Interventions professionnelles, structurelles et organisationnelles en soins primaires visant à réduire les erreurs de médication
Referencias
References to studies included in this review
References to studies excluded from this review
Additional references
References to other published versions of this review
Characteristics of studies
Characteristics of included studies [ordered by study ID]
Ir a:
| Study characteristics | ||
| Methods | Cluster‐RT (randomisation at the pharmacy level) Study duration: 1 year | |
| Participants | 735 at beginning of the study, 600 at the end of the study (data are on the 600 reported below) Setting: community pharmacies Diagnostic criteria: CHD Age (years) (mean): intervention group: 64.8 years; control group: 65.8 years Sex female n (%): intervention group: 79 (29.5%); control group: 94 (29%). Country: Spain Comorbidity: not reported Sociodemographics: not reported Ethnicity: not reported Date of study: not reported | |
| Interventions | 1 intervention group Intervention group: pharmacies allocated to that group provided pharmaceutical care, consisting of the prevention, identification and solution of medication‐related problems. Control group: care as usual Pharmaceutical care consisted of the following: offering the pharmaceutical care service to participants and to their corresponding GPs, initial interview and assessment of the therapeutic plan, registration of data during the subsequent visits in order to allow the identification of medication‐related problems, and intervention to solve the problem. | |
| Outcomes |
| |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Not reported |
| Allocation concealment (selection bias) | Unclear risk | Not reported |
| Blinding of participants and personnel (performance bias) | Unclear risk | Not reported. Pharmacists may have been blinded, as they all received training on methods to treat CHD (in order to ensure that differences after the intervention are due to the intervention per se and not due to differences in theoretical knowledge on methods to treat CHD). |
| Incomplete outcome data (attrition bias) | High risk | High proportion of incomplete outcome data for most to the measures |
| Blinding of outcome assessment (detection bias) | Unclear risk | Not reported |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported |
| Protection against contamination bias | Unclear risk | Unclear |
| Other bias | High risk | For most of the outcome measures no pre‐intervention data were collected. We considered other bias due to cluster randomisation. |
| Study characteristics | ||
| Methods | RT Study duration: 18 months Setting: community pharmacy | |
| Participants | 2454 participants were recruited: 1290 intervention participants and 1164 control participants were assessed at baseline although there were subsequent dropouts. Diagnostic criteria: participants were eligible if they were ≥ 65 years, taking 4 or more prescribed medicines, and oriented with respect to time, place, and person. They were required to be community dwelling and regular visitors to a community pharmacy. Participants could not be housebound or in a nursing facility. Age (years) (mean ± SD): intervention: 735 (58%); control: 663 (57%); no significant difference Sex female n (%): intervention: 735 (58%); control: 663 (57%); no significant difference Country: 7 European countries; Denmark, Germany, The Netherlands, Northern Ireland (co‐ordinating centre), Portugal, Republic of Ireland and Sweden. Comorbidity: there were no significant differences between intervention and control participants at baseline. Sociodemographics: none of note, although participants from 2 countries (Republic of Ireland and Portugal) did not complete the study Ethnicity: not reported Date of study: unclear although published in 2001 | |
| Interventions | 104 intervention pharmacies A pharmaceutical care programme was involved and a manual was distributed to all the intervention sites detailing the intervention. Pharmacists assessed participants to identify drug‐related problems using a structured approach. Pharmacists used several sources of information including informal questioning of the participant, the participant's GP, and pharmacy records. Pharmacists also formulated a monitoring and intervention plan for each participant, which included participant education about drugs and their medical condition, using improvement in medication compliance strategies, and simplifying drug regimens. Control group: participants were treated as per the usual care with no pharmaceutical care plan provided. | |
| Outcomes | Data relating to health and economic outcomes were collected for each participant at baseline, 6, 12 and 18 months. These included hospital admissions. | |
| Notes | The study authors note that the training of pharmacists was not rigorously controlled. Although a study manual was provided along with a 1‐day training session, additional training was not consistently provided. | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Sites were randomly allocated as control or intervention sites |
| Allocation concealment (selection bias) | Low risk | Concealment was adequate as sites rather than individual participants were randomly allocated. Also, all units were allocated at the start of the study. |
| Blinding of participants and personnel (performance bias) | Low risk | As pharmacies were the unit of randomisation, this appears to be low risk. Control pharmacists provided usual care and intervention pharmacists only provided the intervention. |
| Incomplete outcome data (attrition bias) | High risk | Those participants who withdrew from the study were significantly older and in poorer health. |
| Blinding of outcome assessment (detection bias) | Low risk | This is less important for our purpose as we are looking at objective outcomes (hospitalisations, ED visits, and mortality). |
| Selective reporting (reporting bias) | Low risk | No evidence of selective reporting |
| Protection against contamination bias | Low risk | No evidence of contamination of the intervention group with the control |
| Other bias | Unclear risk | 1 aspect of the study that was not rigorously controlled was the training of participating pharmacists. A study manual was provided to each participating pharmacist, followed by a 1‐day training session. Further training was provided in individual countries; however, the extent of this was driven by the available resources. |
| Study characteristics | ||
| Methods | RT 15 months | |
| Participants | 503 participants: 252 intervention, 251 control; final sample 242 intervention, 246 control Setting: primary care centres Diagnostic criteria: elderly people (> 70 years) on ≥ 8 drugs | |
| Interventions | The intervention consisted of 3 consecutive phases. First, a trained and experienced clinical pharmacist evaluated all drugs prescribed to each participant using the GP‐GP algorithm and based their decision about appropriateness on the STOPP/START criteria. Second, the pharmacist discussed recommendations for each drug with the participant’s physician in order to come up with a final set of recommendations. Finally, these recommendations were discussed with the participant, and a final decision was agreed by physicians and their patients in a face‐to‐face visit. Control group participants followed the usual treatments and control procedures of their physicians. | |
| Outcomes | Main outcome measures regarding intervention effectiveness were as follows
Adherence was measured using the Morisky‐Green test. | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Participants were blindly randomised to 1 or other of the 2 study arms. Assignment was based on a list of random numbers generated by a statistical programme. |
| Allocation concealment (selection bias) | Low risk | Each family physician received 10 sealed, opaque envelopes with identification numbers (assigned consecutively in strict chronological order of recruitment) on the back. Each envelope contained a card with the same identification number and the intervention group to which the subject was assigned. Envelopes were not prepared in primary care centres but in the research unit. |
| Blinding of participants and personnel (performance bias) | Low risk | Low risk because pharmacists only treated intervention participants and did not know that the participants they interacted with were in a study. Also, participants did not appear to know whether they were receiving an intervention or not. |
| Incomplete outcome data (attrition bias) | Low risk | Very few people dropped out. |
| Blinding of outcome assessment (detection bias) | Low risk | Outcomes of interest were objective |
| Selective reporting (reporting bias) | Low risk | No evidence of selective reporting |
| Protection against contamination bias | Unclear risk | Prescribing physicians who received recommendations from the pharmacist regarding intervention group participants also had participants in the control group, so the control group could have benefited from the intervention. |
| Other bias | Unclear risk | Study has limited statistical power to detect effects for outcomes of interest |
| Study characteristics | ||
| Methods | Cluster‐RT. The unit of randomisation was the physician practice Study duration: 2 years | |
| Participants | Total participants: 169 participants, 9 physician groups Participants aged ≥ 65 in ambulatory setting, chronic‐care clinics Age: intervention 77.3%; control 77.4%; no SD provided; P = 0.70 Sex female (%): intervention 47.9%; control 49.6%; no SD provided; P = 0.81 Country: USA Diabetes: intervention 53.2%; control 48.6%; P = 0.62 Education > 12 years: intervention 77.1%; control 66.7%; P = 0.1 Married: intervention 55.2%; control 58.3%; P = 0.63 Income < USD 15,000: intervention 15.8%; control 14%; P = 0.75 Hospitalised in prior year: intervention 46.7%; control 39.7%; P = 0.15 Mean chronic disease score: intervention 7.3; control 7.7; P = 0.06 Mean risk score: intervention 0.55; control 0.53; P = 0.35 Ethnicity: non‐white: intervention 2.8%; control 4.1%; P = 0.54 | |
| Interventions | Intervention practices (5 physicians, 96 participants) held half‐day, chronic‐care clinics every 3‐4 months. These clinics included an extended visit with the physician and nurse dedicated to planning chronic‐disease management, a pharmacist visit that emphasised reduction of polypharmacy and high‐risk medications, and a patient self‐management group. Control practices (4 physicians, 73 participants) received usual care. | |
| Outcomes | Emergency visits (mean/year) and hospitalisations | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Randomisation was not mentioned nor how it was done. |
| Allocation concealment (selection bias) | Unclear risk | Allocation was not mentioned nor how it was done. |
| Blinding of participants and personnel (performance bias) | Low risk | This would be impossible. |
| Incomplete outcome data (attrition bias) | Unclear risk | Some participants did not complete the study and were reported in the study |
| Blinding of outcome assessment (detection bias) | Unclear risk | No explanation given |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported. |
| Protection against contamination bias | Low risk | It is hard to determine if the intervention group interacted with the control group. |
| Other bias | Unclear risk | Unclear. We considered other bias due to cluster randomisation. |
| Study characteristics | ||
| Methods | RT 18 months Chronic‐care geriatric facility | |
| Participants | 359 participants: 183 intervention, 176 control; final sample: 160 intervention, 146 control | |
| Interventions | The intervention consisted of a medication review by the study pharmacist for all residents at study opening and 6 and 12 months later using the The STOPP/START criteria. Interventional recommendations that the study pharmacist made for residents in the intervention group but not in the control group were discussed with the chief physician at study opening and after 6 months. The chief physician decided whether to accept these recommendations and implement prescribing changes. Control: usual care | |
| Outcomes | Outcome measures included:
Outcomes were measured at the beginning of the study and at the 12‐month follow‐up. | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Fixed, stratified randomisation was used to allocate residents to groups according to the 3 types of residents: ADL‐dependent, ADL‐independent, and primarily cognitively impaired. Participants who were ADL‐dependent with impaired cognition were assigned to the ADL‐dependent group. Randomisation for each level was according to simple list randomisations. |
| Allocation concealment (selection bias) | Low risk | A physician who was not part of the study randomised participants. Group allocation was concealed from the study pharmacist, and participants were assigned to 1 of the 2 groups using sealed envelopes. |
| Blinding of participants and personnel (performance bias) | High risk | Pharmacists were aware that they were interacting with the intervention group. |
| Incomplete outcome data (attrition bias) | Low risk | Most participants completed the study and there was no apparent difference between intervention and control group. |
| Blinding of outcome assessment (detection bias) | Low risk | Number of hospitalisations is an objective measure. |
| Selective reporting (reporting bias) | Low risk | No evidence of selective reporting |
| Protection against contamination bias | Unclear risk | The intervention pharmacist only interacted with intervention participants, but may have interacted with pharmacists in the control group |
| Other bias | High risk | Only 1 geriatric centre was investigated |
| Study characteristics | ||
| Methods | RT | |
| Participants | 1018 residents: 516 intervention, 502 control Final sample: 59 physicians, 716 nursing home residents Intervention: 29 doctors, 372 nursing home residents Control: 30 doctors, 344 nursing home residents Diagnostic criteria: residents aged ≥ 65 years and clinically stable (no change in prescription in last 2 months) Setting: private organisation | |
| Interventions | 6 months professional intervention A nursing home physician, expert in drug use in older people, delivered a structured educational intervention. The programme included: general aspects of prescription and drug use in geriatric patients, how to reduce the number of drugs, to perform a regular review of medications, to avoid inappropriate drug use, to discontinue drugs that do not show benefits, and to avoid under‐treatment with drugs that have shown benefits. It also discussed in detail some drugs frequently related to adverse drug reactions in older people. Educational material and references were given to participants. Finally, two, 1‐h workshops reviewed practical, real life cases and promoted practice changes in participants. The educator offered further on‐demand advice on prescriptions for the next 6 months. This intervention was reinforced through a single review by the researchers, using standard appropriateness criteria, STOPP‐START. Control: physicians in the control group did not receive any intervention or information about an educational intervention delivered in other centres. | |
| Outcomes | Outcome measures were as follows:
| |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Randomisation was done using random number tables. |
| Allocation concealment (selection bias) | Unclear risk | There was no mention made of sequence concealment. |
| Blinding of participants and personnel (performance bias) | Low risk | Physicians in both groups were informed that there was a company programme aimed to improve drug prescription (to explain why data on prescriptions was collected in their centres) but were blinded to the fact that the educational intervention was being assessed. Also, participants did not know they were receiving an intervention. |
| Incomplete outcome data (attrition bias) | Unclear risk | 30% of participants were lost to the study, but it is unclear if there was differential attrition in intervention and control groups. |
| Blinding of outcome assessment (detection bias) | Low risk | Number of emergency room visits and length of hospitalisations are objective outcomes. |
| Selective reporting (reporting bias) | Low risk | No evidence of selective reporting |
| Protection against contamination bias | High risk | Although nursing homes in the intervention and control groups were separate, some cross‐contamination because of informal contacts between physicians may have occurred. |
| Other bias | High risk | Short intervention period (6 months) and short follow‐up (3 months) |
| Study characteristics | ||
| Methods | Cluster‐RT | |
| Participants | 656 home care participants (intervention n = 297, usual care n = 359) were available for this study | |
| Interventions | The intervention began approximately 3 days after in‐home health admission when a pharmacy technician completed telephonic medication reconciliation with the participant and/or caregiver. Then a trained pharmacist would consult with the participant or caregiver via telephone for an average of 30 min to complete a scheduled comprehensive medication therapy review to identify and resolve any medication‐related problems. The pharmacist constructed a personal medication record and a medication‐related action plan for the participant. The action plan was a participant‐centred document that assisted participants, caregivers, and the pharmacist in the resolution of identified medication‐related problems. Control group: standard/usual care | |
| Outcomes | The primary outcome of this study was participant‐level, 60‐day, all‐cause ED utilisation. This outcome was defined as a dichotomous variable (i.e. the participant visited the ED 1 or more times following the intervention or they did not). | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Site and participant randomisation was a 2‐step process. Firstly a simple random sample of 40 co‐ordinating home healthcare centres, with a monthly census of ≥ 20 admitted participants, was selected among 419 care centres from a nationwide Home Health Agency (Amedisys, Inc, Baton Rouge, LA). Then, at each study site, using blocks of 7 participants, and constrained for equal allocation to study intervention or usual care groups. |
| Allocation concealment (selection bias) | Low risk | Home Health Agency nurses were blinded to their participants’ group assignment to prevent bias during the initial in‐home admissions assessment. |
| Blinding of participants and personnel (performance bias) | Low risk | Participants did not know to which group they were allocated. |
| Incomplete outcome data (attrition bias) | Low risk | All participants were accounted for. |
| Blinding of outcome assessment (detection bias) | Unclear risk | Not mentioned |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported. |
| Protection against contamination bias | Low risk | Cluster‐randomisation with a small chance of contamination |
| Other bias | Unclear risk | Unclear |
| Study characteristics | ||
| Methods | RT, total study duration not provided | |
| Participants | 5077 hospital discharges: 2563 intervention discharges, 2514 control discharges Final sample: 1870 intervention, 1791 control Setting: large multispecialty group practice Diagnostic criteria: ≥ 65 years discharged from hospital to home | |
| Interventions | Professional intervention Intervention: an automated system was developed to facilitate the flow of information to the medical group’s primary care providers about individuals who were discharged to home from the hospital. In addition to notifying providers about an individual’s discharge, the system provided information about new drugs at the time of hospital discharge, warnings about selected drug‐drug interactions, recommendations for consideration of dose changes and laboratory monitoring of high‐risk medications, and alerts to the provider’s support staff to schedule a post hospitalisation office visit within 1 week of discharge. Control: care as usual | |
| Outcomes | Whether discharged individuals had an office visit with a primary care physician in the 7‐, 14‐, and 30‐day periods after hospital discharge was determined, as was whether a participant was rehospitalisation within 30 days. Information related to office visits and hospitalisations was ascertained from the medical group’s electronic health record and from health plan data, which allowed for determination of whether a rehospitalisation had occurred at any hospital and not just the primary hospital that served individuals under the care of the medical group. Analysts blinded to intervention status determined these outcomes at least 6 months after completion of the study. | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | A random number generator was used to assign a discharge to the intervention or control group. |
| Allocation concealment (selection bias) | Low risk | Computer allocated discharges |
| Blinding of participants and personnel (performance bias) | Low risk | Automated system was used. Also participants were not aware of which group they were in |
| Incomplete outcome data (attrition bias) | Low risk | Comparable rates of attrition in intervention and control groups. |
| Blinding of outcome assessment (detection bias) | Low risk | Analysts blinded to intervention status determined the outcomes. (The trialist reviewing the data (JHG) was unaware of which type of unit the event had occurred on.) |
| Selective reporting (reporting bias) | Low risk | No evidence of selective reporting. |
| Protection against contamination bias | Unclear risk | The study is a cluster design and the authors stated the following, "Efforts were made to limit crossover of prescribers between intervention and control units, however, some prescribers worked simultaneously on both intervention and control units." |
| Other bias | Unclear risk | Unclear risk |
| Study characteristics | ||
| Methods | RT 18 months | |
| Participants | 61 participants: 24 intervention, 37 control Unclear how many participants were analysed Setting: healthcare system's outpatient family medicine centre Diagnostic criteria: In the first year of the study, inclusion criteria had to be 1 of the following 3 criteria:
In the second year, the criteria were changed to the following: ≥ 8 scheduled medications anticipated at discharge | |
| Interventions | Organisational. Participants in the intervention group were scheduled for a care transitions clinic visit with a clinical pharmacist approximately 72 h post discharge, and prior to the post‐hospitalisation, primary care‐provider visit. The visit involved performing a complete medication history, identifying and resolving medication discrepancies, creating a current medication list for both the medical record and the participant, and counselling on appropriate medication use. During these visits, the pharmacist identified discrepancies between the best possible medication discharge list and the discharge summary, and characterised medication discrepancies using predefined categories. Study participants in the usual care group were scheduled to see their primary care provider for a post‐hospitalisation visit with no interim pharmacist intervention. Medication discrepancies of study participants not attending care transitions visits were identified and characterised by study personnel in the same manner as those in the intervention group. Study personnel reviewed study participants’ medical records to quantify 30‐day ED visits and rehospitalisation at the study institution. All study participants received a phone call approximately 30 days after discharge to report hospitalisations or ED visits at outside institutions. Only hospitalisations and ED visits at the study institution were included for those participants who were not able to be contacted after 3 phone call attempts. Both the intervention and control group received clinical pharmacy services for the family medicine inpatient service and outpatient family medicine clinic. Inpatient clinical pharmacists conducted rounds with the medical team daily, reviewed and monitored medications for effectiveness and safety, and made recommendations to the physician staff to optimise medications. Participants in both groups received this usual care from the inpatient pharmacist. The role of the inpatient pharmacist in the study was to collaborate with the inpatient medical team to create a BPMDL for all study participants just prior to discharge. The BPMDL was used to identify medication discrepancies, and it served as the gold standard list of medications that the participant should take after discharge. The BPMDL accounted for home medications, medication changes made during the hospitalisation, and medications that should be initiated or discontinued on discharge. | |
| Outcomes | The 3 prespecified primary outcomes of this study were a composite of the occurrence of a hospital admission or an ED visit within 30 days after hospital discharge and the resolution of medication discrepancies before the primary care provider visit. Secondary outcomes include the individual rates of rehospitalisation and ED visits within 30 days after discharge. | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | For the first year of study, a random number generator was used to randomise participants. For the second year, block randomisation with a block size of 4 was used. |
| Allocation concealment (selection bias) | Low risk | A computer was used to randomise |
| Blinding of participants and personnel (performance bias) | Unclear risk | Participants did not know they were receiving the intervention, however, pharmacists may have been aware that they were delivering the intervention. |
| Incomplete outcome data (attrition bias) | High risk | Half of the intervention participants did not participate in the clinic visit. |
| Blinding of outcome assessment (detection bias) | Low risk | Hospital admissions and emergency room visits are objective outcomes. |
| Selective reporting (reporting bias) | Low risk | There was no evidence of selective reporting. |
| Protection against contamination bias | Unclear risk | It cannot be determined if the control group was contaminated by the intervention. |
| Other bias | Unclear risk | Small sample size. Also, recall bias may have operated as participants provided a self‐report of hospitalisations and ED visits outside of the study institution. It is not clear what was the final analysis sample. |
| Study characteristics | ||
| Methods | RT Setting: home visit study (primary care) Study duration: 6 months | |
| Participants | 872 participants in a home visit study. Researchers recruited patients from four general hospitals and six community hospitals if they were aged ≥ 80 years, admitted as an emergency, intended to be discharged to their own home or warden‐controlled accommodation, and prescribed ≥ 2 drugs on discharge. Exclusion criteria: participants received dialysis treatment and participation in an intensive discharge service on 1 site Age (years) (mean ± SD): intervention 85.4 (4); control 85.5 (4); not significant Sex female n (%): intervention 262 (61.1); control 272 (63.8); not tested for significance Country: UK Comorbidity: baseline diagnosis: cardiovascular (total): intervention 134 (31.2), control 144 (33.8); myocardial infarction/angina: intervention 57 (13.3), control 65 (15.3); heart failure: intervention 38 (8.9), control 34 (8.0); musculoskeletal (total): intervention 61 (14.2), control 65 (15.3); fracture: intervention 37 (8.6), control 40 (9.4); gastrointestinal (total): intervention 47 (11.0), control 54 (12.7); respiratory (total): intervention 48 (11.2), control 49 (11.5); COPD/asthma: intervention 15 (3.5), control 13 (3.1); lower respiratory tract infection: intervention 16 (3.7), control 22 (5.2); neurological: intervention 40 (9.3), control 25 (5.9); stroke/transient ischaemic attack: intervention 16 (3.7), control 14 (3.3); senility/dementia: intervention 16 (3.7), control 6 (1.4); genitourinary: intervention 17 (4.0), control 16 (3.8); cancer (total): intervention 15 (3.5), control 7 (1.6); other or unclassified: intervention 67 (15.6), Control 66 (15.5) Sociodemographic: not mentioned Ethnicity: not mentioned Participants recruited between October 2000 and December 2002 | |
| Interventions | Initial referral to a review pharmacist included a copy of the participant's discharge letter. Pharmacists arranged home visits at times when they could meet participants and carers. Pharmacists assessed participants' ability to self‐medicate and drug adherence, and they completed a standardised visit form. Where appropriate, they educated the participant and carer, removed out of date drugs, reported possible drug reactions or interactions to the general practitioner, and reported the need for a compliance aid to the local pharmacist. Where a compliance aid was recommended, this was provided within the trial and a filling fee was paid to the local pharmacist. 1 follow‐up visit occurred at 6‐8 weeks after recruitment to reinforce the original advice. Control participants received usual care. | |
| Outcomes | The primary outcome was total number of emergency admissions to hospital over 6 months. Secondary outcomes included deaths, admissions to residential homes and nursing homes, and self‐assessed QoL measured using the EQ‐5D. Participants also rated their health on a visual analogue scale from 100 (perfect health) to 0 (worst imaginable health). The EQ‐5D and visual analogue scales were collected at baseline, 3 months, and 6 months. Data were collected on emergency admissions from hospital episode statistics. | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Third party telephone randomisation based on a computer‐generated sequence in blocks of varying length. Randomisation appeared adequate |
| Allocation concealment (selection bias) | Low risk | Sequence was concealed based on what is noted above about sequence generation |
| Blinding of participants and personnel (performance bias) | High risk | This was not done as stated in the manuscript, "Because of the nature of the intervention, no “placebo” could be provided. Participants were told after randomisation which group they were in." |
| Incomplete outcome data (attrition bias) | Low risk | Follow‐up of the main outcome (hospital admissions) was good—only 3% of participants withdrew or were lost to follow‐up. |
| Blinding of outcome assessment (detection bias) | Low risk | This is less important for our purpose as we are looking at objective outcomes (hospitalisations, ED visits, and mortality). |
| Selective reporting (reporting bias) | Low risk | No evidence of selective reporting |
| Protection against contamination bias | Unclear risk | It can not be determined if the control group has been contaminated by intervention |
| Other bias | Low risk | There is no other bias. |
| Study characteristics | ||
| Methods | RT Setting: community participants Follow up: 3 months | |
| Participants | 240 participants discharged to the community for the first time on oral anticoagulant warfarin (regardless of strength, gender, or age) There was no significant difference between the groups in terms of age, all participants 60.2 years ± 17.84 Sex of participants: male 160 (58.3%). There was no statistically significant difference found between the groups based on indication for anticoagulation, with atrial fibrillation representing the most common indication. All participants lived close to the participating medical centre and could access it easily. Country: United Arab Emirates Comorbidity: atrial fibrillation: 82 (34.2%), valve replacement: 37(15.4%), CHF:32 (13.3%), peripheral artery disease: 8 (3.33%), left ventricular thrombus: 7 (2.91%), stroke: 9 (3.75%) | |
| Interventions | Intervention (Group A) was the 'counselled' group, whereas, control (Group B) was the 'non‐counselled' group. After initial physician/pharmacist consultation in a standard care setting, 1 group was thoroughly counselled, defined by the following:
The other group received no follow‐up consultation other than what was ordered by their own physician in a standard care setting. This group was asked only to visit the anticoagulation clinic twice a month for 3 months to evaluate international normalised ratio levels. | |
| Outcomes | Number of adverse events, emergency visits and inpatients admissions | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Randomisation is not described |
| Allocation concealment (selection bias) | Unclear risk | The first 240 participants discharged or prescribed for the first time warfarin (regardless of strength, gender, or age) were divided randomly and assigned a intervention or control group. |
| Blinding of participants and personnel (performance bias) | Unclear risk | Not fully described: separation of intervention and control groups is not exclusive. |
| Incomplete outcome data (attrition bias) | Low risk | No loss of participants from the study |
| Blinding of outcome assessment (detection bias) | Unclear risk | Not described how and who measured the outcomes |
| Selective reporting (reporting bias) | Low risk | The study reported the proposed outcome. |
| Protection against contamination bias | Unclear risk | It is not clear if the control group was contaminated by the intervention. |
| Other bias | Unclear risk | Not applicable |
| Study characteristics | ||
| Methods | 2‐arm, community, cluster‐randomised trial Study duration: 3 years 39 eligible communities were stratified, geographically defined according to municipal boundaries, by population size (3 strata) and geographical location (4 strata). An independent expert in cluster‐randomised trials then used a random number generator to randomly allocate communities in each stratum to receive either CHAP or no intervention. | |
| Participants | 39 communities (148,589 participants) initially and 145,441 participants after follow‐up post intervention Setting: 39 eligible communities, geographically defined according to municipal boundaries, by population size (3 strata) and geographical location (4 strata); community‐based Sex male (%): intervention communities 42.65 ± 1.19, control communities 42.92 ± 2.16 Country: Ontario, Canada Comorbidity: No. of prescription drugs in previous year: control communities: 7.25 (0.49), intervention communities: 6.98 (0.54) No. of comorbidity groups in previous 2 years: control communities: 7.31 (0.30), intervention communities: 7.17 (0.50) Charlson comorbidity index in previous 2 years: control communities: 0.57 (0.09), intervention communities: 0.58 (0.11) Diabetes (%): control communities: 22.16 (2.34), intervention communities: 21.20 (2.79) History of congestive heart failure (%): control communities: 12.19 (1.91), intervention communities: 12.45 (2.34). Rurality index: control communities 28.96 (13.60), intervention communities: 31.63 (14.09) Low‐income status (%): control communities: 16.95 (8.55), intervention communities: 18.57 (11.33) Ethnicity: Not stated. | |
| Interventions | Communities were randomised to receive CHAP (n = 20) or no intervention (n = 19) In CHAP communities, residents aged ≥ 65 were invited to attend volunteer‐run cardiovascular risk assessment and education sessions held in community‐based pharmacies over a 10‐week period; automated blood pressure readings and self‐reported risk factor data were collected and shared with participants and their family physicians and pharmacists. In both intervention and control arms, residents received the usual health promotion and healthcare services available to all Ontarians under its publicly financed universal health insurance programme. | |
| Outcomes | Rates of hospital admission for acute myocardial infarction, CHF, and stroke in these 39 communities and the 2001 census population estimates for people aged ≥ 65 years and over for power calculations | |
| Notes | A potential limitation of CHAP is the short duration of the intervention. The 10‐week exposure to CHAP may be too short to affect hospital admission rates. | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | The 39 eligible communities were stratified by size of the population ≥ 65 years (3 groups) and geographic location (4 groups), forming seven substrata. Communities within each stratum were randomly allocated to either the intervention (n = 20) or control arm of the study (n = 19) by an independent expert in cluster‐randomised trials not associated with the study. |
| Allocation concealment (selection bias) | Low risk | The independent expert in cluster‐randomised trials was not associated with the study. |
| Blinding of participants and personnel (performance bias) | Low risk | Although intervention community members (older adults, family physicians, volunteers, pharmacists) were clearly aware of their group assignment, the names of control communities were not publicised and control community members were not notified that the study was taking place. |
| Incomplete outcome data (attrition bias) | Low risk | No attrition reported. Cluster‐randomised trial of communities reporting rates of hospitalisation. |
| Blinding of outcome assessment (detection bias) | Low risk | Though unlikely, community members' knowledge of the evaluation could influence outcomes, but hospitalisation rates were retrieved from a population‐based administrative health dataset. |
| Selective reporting (reporting bias) | Low risk | Data retrieved from routinely collected, population‐based administrative health data |
| Protection against contamination bias | Unclear risk | It is hard to determine if the control group is contaminated with the intervention |
| Other bias | High risk | Short duration of the intervention may have had an impact on detection of outcomes such as hospital admissions. We considered other bias due to cluster randomisation |
| Study characteristics | ||
| Methods | RT Setting: ambulatory setting 9 months | |
| Participants | 70 participants Ambulatory setting Attendees at a heart failure outpatient clinic, > 18 years, had New York Heart Association class II, III or IV heart failure, stable signs and symptoms of heart failure, clinically euvolaemic, daily frusemide dose up to a maximum of 320 mg, treatment with other drugs such as beta‐blockers, digoxin, vasodilators and spironolactone was permitted. Participants were excluded if they were not on frusemide; were on a daily frusemide dose above 320 mg and/or thiazide diuretic; had baseline renal impairment (serum creatinine concentration > 200 µmol/L or on dialysis); had a severe psychiatric illness or moderate‐severe dementia; life expectancy of < 3 months; severe hearing impairment or legal blindness; or had difficulty understanding and speaking English and did not have an interpreter or family member to assist. Other exclusions included scheduled cardiac surgery; heart transplant candidacy; inability to give informed consent; and no access to a telephone. | |
| Interventions | Pharmacist intervention focused on participants improving self‐care, recognising symptoms of fluid retention, measuring weight daily and self‐adjusting diuretic dose using frusemide. Intervention group: participants assigned to the intervention group received usual care plus pharmacist intervention. The intervention was provided to every participant in the intervention group and consisted of a 30‐min educational session during the clinic appointment. The pharmacist intervention focused on participants improving self‐care, recognising symptoms of fluid retention, measuring weight daily and self‐adjusting diuretic dose using a flexible frusemide dose‐adjustment regimen, and improving knowledge and understanding of heart failure and heart failure medications. Usual care (control group): usual care was provided to all of the eligible participants by a cardiologist, heart failure nurse co‐ordinators and a dietitian during the clinic appointment. Usual care consisted of assessment of clinical status and medications, education on daily weight measurement, diet, fluid and sodium management, and recognition of signs and symptoms of fluid retention and dehydration. In case of a sudden increase in weight of more than 1 kg/d for 2 d, participants were encouraged to contact the heart failure nurse co‐ordinators for advice in consultation with the cardiologist to self‐adjust their frusemide dose. The heart failure nurse co‐ordinators followed up participants 48 h after a dose adjustment to assess if their weight had decreased and condition improved. The key difference between the groups was that the control group called a heart failure nurse co‐ordinator to discuss frusemide dose modification, while the intervention group adjusted the diuretic dose themselves. | |
| Outcomes | Hospital readmissions due to fluid overload: measured at 1st, 2nd and 3rd months | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Randomisation occurred after selection criteria had been observed. No description of the randomisation method/sequence generation was presented. |
| Allocation concealment (selection bias) | Unclear risk | A significant number of heart failure participants were not good candidates for the intervention. Only 1 in 3 participants who met inclusion criteria remained eligible after application of exclusion criteria. |
| Blinding of participants and personnel (performance bias) | Unclear risk | It is unclear if participants were blinded to the intervention. |
| Incomplete outcome data (attrition bias) | Low risk | All participants completed the study. |
| Blinding of outcome assessment (detection bias) | Low risk | Assessments and readmissions were evaluated and confirmed by an independent doctor blinded to the randomisation using data from participants, hospital admissions records and medical records. |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported as mentioned at the start of the trial. |
| Protection against contamination bias | Unclear risk | It can not be determined if the intervention group interacted with the control group. |
| Other bias | Unclear risk | The intervention was delivered by the same pharmacist. It precludes study of other factors, such as pharmacist attitudes or behaviours that may have promoted delivery of the intervention and limit the generalisability of the intervention. This study was conducted at a single institution, and the results may reflect local population characteristics and patterns of care. |
| Study characteristics | ||
| Methods | RT Duration of study: 3 months | |
| Participants | 332 participants completed study (168 intervention and 164 control) Setting: general practice Diagnostic criteria: the inclusion criteria for participants were ≥ 65 years, regular request for ≥ 4 medicines via the computerised repeat prescribing system and ≥ 2 chronic diseases. Exclusion criteria were dementia and being considered by the GP to be unable to cope with the study. Age (years) (mean ± SD): intervention 74.8 (6.2), control 75.2 (6.6) Sex female n (%): intervention 95 (56.5%), control 106 (64.6%) Country: UK Comorbidity: mean no. of chronic diseases: intervention 3.9 (1.4), control 3.8 (1.4), P = 0.968 Sociodemographics: nothing of note Ethnicity: not mentioned Date of study: not mentioned but paper received by journal on 23 December 1999 | |
| Interventions | 1 intervention group Intervention group: a pharmaceutical care plan was drawn up for each intervention group participant, listing all potential and actual pharmaceutical care issues, together with the desired output(s), the action(s) planned to achieve the output(s) and the outcomes of any potential pharmaceutical care issues already resolved by the pharmacist. Copies of the plan were inserted in the participants' medical notes and given to their GP, who was asked to indicate their level of agreement with each pharmaceutical care issue identified and with the actions. The pharmacist then implemented all remaining agreed actions, assisted by other practice staff where appropriate. Control participants were similarly interviewed and pharmaceutical care issues identified, although no pharmaceutical care plan was implemented. Participants were advised to consult any usual carers or health‐care professionals in response to direct queries during interview. | |
| Outcomes | Number of hospital admissions | |
| Notes | The pharmacists undertaking the medication review also administered the SF‐36 questionnaire and identified all care issues. There is also potential for GPs receiving recommendations for some participants to increase their tendency to note similar issues in control participants. In some cases the care plan was not fully implemented by 3 months. | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Using random number tables, 1 practice from each of the 6 resultant categories was selected and invited to participate. 1 practice refused and a further practice was randomly selected. Participants were randomly allocated to the intervention or control group. |
| Allocation concealment (selection bias) | Unclear risk | Although a random number table was used to select practices, it was not clear whether participant assignment to intervention and control groups was concealed. |
| Blinding of participants and personnel (performance bias) | Low risk | This was adequate because pharmacists only treated intervention participants and did not know that the participants they interacted with were in a study. Also, participants did not appear to know whether they were receiving an intervention or not. |
| Incomplete outcome data (attrition bias) | Low risk | Equal numbers of participants in the control and intervention groups withdrew from the study. Around 14% to 15 % of the participants withdrew in each group. |
| Blinding of outcome assessment (detection bias) | Low risk | This is less important for our purpose as we are looking at objective outcomes (hospitalisations, ER visits, and mortality) |
| Selective reporting (reporting bias) | Low risk | There is no evidence that there was selective reporting of results |
| Protection against contamination bias | High risk | It can not be determined if the control group was contaminated with the intervention group. |
| Other bias | Unclear risk | In some cases, the care plan had not been fully implemented by the 3‐month follow‐up. |
| Study characteristics | ||
| Methods | Cluster‐RT Study duration: 2 years | |
| Participants | 6523 participants Final sample: 1769 control, 1769 intervention Diagnostic criteria: not relevant as homes were the unit of analysis not individuals Age (years) (mean ± SD): average age of residents was not reported. At baseline 16% of the residents in both intervention and control homes were aged 65‐74 years, 36% in intervention homes and 35% in usual care homes were 75‐84 years, and 40% of the residents in the intervention homes and 36% in the usual care homes were ≥ 85 years. During the intervention period, 15% in both groups were 65‐74 years, 39% were 75‐84 years, and 39% were ≥ 85 years Sex, female n (%): at baseline, 72% of the residents in the intervention homes and 68% in the usual‐care homes were female. During the intervention period, 74% of the residents in the intervention and usual‐care homes were female. Country: USA Comorbidity: (intervention, control), dementia (35.4, 43.4); Alzheimer's disease (12.7, 14.6), cancer (8.3, 12.1), diabetes mellitus (27.5, 31.0), cerebrovascular accident (22.2, 22.4), heart failure (26.5, 28.5), coronary artery disease (18.6, 16.2), arrhythmia (15.8, 15.8), hypertension (64.9, 61.8), other cardiovascular disease (23.6, 28.0) Sociodemographics: nothing reported other than race Ethnicity: 18% in intervention group and 11% in usual care group were minority race at baseline. During the intervention period, 19% of both groups were minority race Date of study: 2003‐2004 | |
| Interventions | Professional intervention The overarching idea was to use health information technology to engage consultant pharmacists and nursing staff to identify residents at risk for delirium and falls, implement proactive monitoring plans as appropriate, and provide reports to assist consultant pharmacists in conducting the medication regimen review. Intervention: A Geriatric Risk Assessment MedGuide database for falls and delirium was integrated into the pharmacies' commercial pharmacy software system (Rescot LTCP System) for the intervention homes. Control: usual care | |
| Outcomes | Incidence of potential delirium, falls, hospitalisations potentially due to adverse drug events, and mortality | |
| Notes | Residents in the intervention homes experienced fewer falls, less potential delirium, and death, but more hospitalisations than in the comparison homes. In new admissions, there appeared to be a trend toward lower mortality (adjusted hazard ratio 0.88, 95% CI 0.66 to 1.16) and a lower overall hospitalisation rate (adjusted hazard ratio 0.89, 95% CI 0.72 to 1.09) and a clear reduction in the rate of potential delirium (adjusted hazard ratio 0.42, 95% CI 0.35 to 0.52) in the intervention homes than the comparison homes. | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Not reported |
| Allocation concealment (selection bias) | Unclear risk | Not reported |
| Blinding of participants and personnel (performance bias) | Low risk | Cluster‐randomisation |
| Incomplete outcome data (attrition bias) | Low risk | All participants were reported. |
| Blinding of outcome assessment (detection bias) | Unclear risk | Not stated |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported. |
| Protection against contamination bias | Low risk | It can not be determined if the intervention group was contaminated with the control group. |
| Other bias | Unclear risk | There is no evidence of the presence of other bias. We considered other bias due to cluster randomisation. |
| Study characteristics | ||
| Methods | RT 6 months | |
| Participants | 136 participants registered with 1 general practice (1 participant from each group withdrew shortly after randomisation) Home‐based > 80 years, living at home, taking ≥ 4 oral medications, and had ≥ 1 additional medicine‐related risk factor Participants were excluded if they were resident in a care home or if there was documented use of an adherence aid. Age: intervention 84.5 years, control 84.1 years (no SD supplied) Gender female: intervention 46 (67.6%), control 42 (63.6%) Country: UK Sociodemographics: living alone: intervention 44 (64.7%), control 43 (65.1%); social class (I, II, III): intervention 33 (48.5%), control 29 (43.9%); 9% of practice were aged over 80 years (twice the national average) Ethnicity: 98.5% of the local town population were white, compared to 90.9% for England | |
| Interventions | Comparing home‐based medication review with standard care The intervention: the pharmacist was asked to identify cases where adverse drug reactions or drug interactions may be occurring. This was noted using a tick box on the medication review form after detailed information had been gained from the participant regarding all over‐the‐counter and prescribed drugs The control group received standard care | |
| Outcomes |
| |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | No indication of random sequencing |
| Allocation concealment (selection bias) | Low risk | Randomisation was carried out by a third party and was stratified by whether the participant lived alone. |
| Blinding of participants and personnel (performance bias) | Unclear risk | It is unclear if participants were blinded to the intervention. |
| Incomplete outcome data (attrition bias) | Low risk | Attrition reported and reasons for attrition presented |
| Blinding of outcome assessment (detection bias) | Unclear risk | Outcome data on hospital admissions were provided by hospital episode statistics (not self‐report) and are therefore unlikely to be biased. |
| Selective reporting (reporting bias) | Low risk | Outcome data on hospital admissions were provided by hospital episode statistics (not self‐report) and are therefore unlikely to be biased. |
| Protection against contamination bias | Unclear risk | Unclear |
| Other bias | High risk | Research was carried out in 1 rural general practice with a single experienced review pharmacist, which has a bearing on the generalisability of the results. |
| Study characteristics | ||
| Methods | Study design: a cluster‐randomised design, this provides protection against contamination across trial groups when trial participants are managed within the same setting. Participants in practices in the UK were managed by all GPs within the practice; the control intervention was mediated by GPs, this precluded individual, participant‐level randomisation. Study duration: median follow‐up was 4.7 years | |
| Participants | 2164 participants (174 practices) Setting: general practice Diagnostic criteria: consenting participants were eligible if aged ≥ 18 years and had left ventricular systolic dysfunction confirmed by cardiac imaging conducted at a local hospital (transthoracic echocardiography in 90% of cases). Participants did not have to have symptoms or signs of heart failure. Family doctors received a semi quantitative report of left ventricular systolic function (normal, mild, moderately or severely reduced) instead of ejection fraction. Age (years) (mean ± SD): pharmacist intervention, 70.6 (10.3) and control 70.6 (10.1) Sex female n (%): pharmacist intervention 320 (29%), control 329 (31%) Country: UK Comorbidity: hypertension, myocardial infarction, pharmaceutical care issue, coronary artery bypass grafting, atrial fibrillation or flutter, diabetes mellitus, stroke, respiratory disease, asthma Sociodemographics: not mentioned Ethnicity: not mentioned Date of study: from 25 October 2004‐6 September 2007 | |
| Interventions | 1 intervention and 1 control Participants from practices assigned to the intervention were offered a 30‐min appointment with a pharmacist. The main aim of this review was optimisation of medical treatment for left ventricular systolic dysfunction according to guidelines (supplementary material online). If there was agreement between the pharmacist and the participant during the consultation and subsequently with the family doctor, medications were initiated, discontinued, or modified by the pharmacist during 3‐4 subsequent weekly or fortnightly consultations. Family doctors provided usual care thereafter. No instructions were given to family doctors in the usual care practices. The study pharmacists did not collect information on symptoms or examine the participants as this was not part of their professional training. | |
| Outcomes |
| |
| Notes | There was no difference in mortality or hospital admissions between the intervention and the control group. (Mortality from heart failure should be reported in the final analysis as this intervention was targeting heart failure management) | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Randomisation was generated by a computer |
| Allocation concealment (selection bias) | Low risk | Computer |
| Blinding of participants and personnel (performance bias) | Low risk | Cluster randomisation was undertaken |
| Incomplete outcome data (attrition bias) | Low risk | All participants were reported |
| Blinding of outcome assessment (detection bias) | Unclear risk | Not stated |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported |
| Protection against contamination bias | Low risk | It can not be determined if the intervention group mixed with the control group |
| Other bias | Unclear risk | Unclear. We considered other bias due to cluster randomisation. |
| Study characteristics | ||
| Methods | Cluster‐RT (pharmacies were the cluster unit of randomisation) | |
| Participants | 31; 17 intervention and 14 control, this was also the final sample that was analysed Setting: community pharmacists Diagnostic criteria: participants were ≥ 65, used ≥ 5 medications for ≥ 6 months, with the ability to complete the EuroQol 5D questionnaire | |
| Interventions | Organisational IIntervention group: pharmacists allocated to the intervention group provided the medication with follow‐up service according to national guidelines. The medication review with follow‐up service started with a comprehensive interview undertaken in a private area of the pharmacy. The pharmacist collected relevant information about the participant’s health problems, medicines used, clinical and biological parameters (gathered through medical records provided by the participant or measured in the pharmacy), medication use, lifestyle habits and concerns about diseases and medications. Pharmacists also assessed the level of control of health problems by using information referred by participants’ and/or clinical and biological parameters, depending on the type of health problem (i.e. pain versus hyperlipidaemia) and classified every health problem as controlled, uncontrolled or unknown. After performing a comprehensive medication review, the pharmacist identified negative clinical outcomes related to medicines and drug‐related problems. Subsequently, an action plan was agreed upon by the participant and the physician if required. This medication review with follow‐up service was focused on both participants’ outcomes and medication use process and required a commitment to follow‐up. The usual care consisted of dispensing medicines prescribed by physicians and advice on minor ailments. | |
| Outcomes | Medication‐related hospital admission was the primary outcome of this sub analysis. Hospital admissions were recorded in participants’ visits to the pharmacies and the medication related ones were identified through the expert panel after the fieldwork. Kappa values ranging from 0.61 to 1 were considered as an acceptable incidence rate ratio to measure the agreement among experts. The cost of hospital admissions estimated by diagnosis‐related group was a secondary outcome and the diagnosis‐related groups were recorded after the fieldwork. | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Pharmacies were randomised to the intervention or control group by an independent researcher. |
| Allocation concealment (selection bias) | Low risk | An independent researcher performed randomisation using a computer‐generated list of random numbers. |
| Blinding of participants and personnel (performance bias) | High risk | Neither the participants nor pharmacists were blinded. |
| Incomplete outcome data (attrition bias) | Low risk | There was little attrition and comparable rates for intervention and control group. |
| Blinding of outcome assessment (detection bias) | Low risk | For the sub analysis, the expert panel was blind as to which group the participants belonged so whether a hospital admission was medication‐related was not affected. |
| Selective reporting (reporting bias) | Low risk | No evidence of selective reporting |
| Protection against contamination bias | Unclear risk | Although pharmacies were randomised to control and intervention groups, informal contact between pharmacists may have led to contamination. |
| Other bias | Unclear risk | There is no evidence of other bias. |
| Study characteristics | ||
| Methods | A prospective, multisite RT Duration of study: 12 months | |
| Participants | Of 1054 participants enrolled at the 9 Veterans Affairs clinics, 523 were randomised to the intervention and 531 to the control group. Of these, 950 participants completed 6‐month follow‐up questionnaires and 931 completed the study. Of participants completing the study, 447 were in the intervention group and 484 were in the control group. Setting: Veterans Affairs clinics Interventions: clinical input by pharmacists Diagnostic criteria: participants were considered at high risk for drug‐related problems if they met ≥ 3 of the following criteria: were taking ≥ 5, were taking ≥ 12 doses/d, had ≥ 3 chronic medical conditions, had ≥ 4 changes in their drug regimen over the past year, had a history of noncompliance with drug therapy, or were taking an agent that required therapeutic drug monitoring. Age: (years) mean ± SD: 67 ± 10.1 Sex n (%): intervention group 21 (0.04%), control, 20 (0.04%) Country: USA Comorbidity: hypertension, angina, hyperlipidaemia, arthritis, diabetes and COPD Sociodemographics: not mentioned Ethnicity: not mentioned | |
| Interventions | 1 intervention and 1 control The intervention group was given a protocol to follow; the protocol indicated that each participant should have ≥ 3 visits with the clinical pharmacist during the study, but participants could be seen as frequently as deemed necessary to ensure appropriate care. Visits were to occur between or concurrent with appointments with the primary care provider or other physicians. The control group followed the usual care with no specific protocol given to clinicians. | |
| Outcomes | Number of hospitalisations | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Participants were randomly assigned using a central computer. |
| Allocation concealment (selection bias) | Low risk | Computer‐based |
| Blinding of participants and personnel (performance bias) | Unclear risk | Not possible |
| Incomplete outcome data (attrition bias) | Low risk | Some participants did not complete the study and were reported in the study. |
| Blinding of outcome assessment (detection bias) | Unclear risk | Unclear how it was done |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported |
| Protection against contamination bias | Unclear risk | It is not possible to determine if the control group was contaminated with the intervention group. |
| Other bias | Unclear risk | It is unclear if there are other biases. |
| Study characteristics | ||
| Methods | Prospective, randomised study design Study duration: 2 years The major limitation of the study was selection bias because only participants who responded to a letter of invitation had the opportunity to take part in the study. | |
| Participants | 96 participants took part in the study; 48 were randomised to the nurse group and 48 to the non‐nurse group. Setting: outpatient heart failure clinic Diagnostic criteria: participants who survived index hospitalisation were invited by letter to a visit for treatment optimisation at the outpatient heart failure unit. Among the participants who appeared at the ambulatory visit, those with a verified heart failure diagnosis and residing < 50 km from Vienna were eligible for the nurse intervention and therefore offered to participate in the present study. Baseline evaluation was performed by a cardiologist specialising in the management of heart failure. The ambulatory visit comprised a patient history, physical examination, electrocardiogram, a routine blood analysis, and, if necessary, an echocardiography. Furthermore, blood samples were taken for later analysis of natriuretic peptides. The participants were thoroughly informed about the disease of CHF and recommendations were made regarding medication, self‐assessment of weight, blood pressure and pulse, and diet and exercise management. The baseline demographic, clinical, and therapeutic characteristics were not statistically different between the nurse group and the non‐nurse group. Age (years) (mean ± SD): non‐nurse, control (66 ± 13); nurse, intervention 70 ± 12 Country: Vienna, Austria Comorbidity: hypertension, diabetes, respiratory diseases Sociodemographics: not reported Ethnicity: Austrian (unclear if they were all white) | |
| Interventions | 1 intervention and 1 control. There were 48 participants in each group. Intervention: home‐based nurse care Participants in the nurse group were visited by a nurse specialised in caring for people with heart failure on the initial visit at the outpatient heart failure unit and then at their home 3, 6, 9, and 12 months after randomisation. At home visits, the nurse checked and recorded weight, symptoms and signs of heart failure, heart rate and blood pressure, and organised and reviewed blood analyses on demand, especially of electrolytes and renal parameters. Furthermore, the nurse had to check for and, in co‐ordination with the treating physician, implement guideline‐based medication. Moreover, the nurse was in charge of individualised participant and caregiver education and enhancement of self‐management. If the nurse noted any deterioration in the participant's status, she reported to the treating physician or advised the participant to visit the treating physician. Control group received the usual care provided | |
| Outcomes | Admission for heart failure at 12 months and 24 months Mortality at 12 months and 24 months | |
| Notes | The major limitation of the study is selection bias because only participants who responded to a letter of invitation had the opportunity to take part in the study. See notes to other relevant studies | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | It is unclear how the sequence generation was done. |
| Allocation concealment (selection bias) | High risk | The major limitation of the study was selection bias because only participants who responded to a letter of invitation had the opportunity to take part in the study. |
| Blinding of participants and personnel (performance bias) | Unclear risk | It is unclear if participants were blinded to the intervention. |
| Incomplete outcome data (attrition bias) | Low risk | There is no incomplete outcome data. |
| Blinding of outcome assessment (detection bias) | Unclear risk | Independent data collector |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported as mentioned at the start of the trial. |
| Protection against contamination bias | Unclear risk | Unclear |
| Other bias | High risk | Small sample size |
| Study characteristics | ||
| Methods | RT with a 2 x 2 factorial design using physician and pharmacist interventions, which resulted in 4 groups of participants: physician intervention only, pharmacist intervention only, intervention by physician and pharmacist, and intervention by neither physician nor pharmacist (control) Study duration: 1 year | |
| Participants | Total participants 712 with uncomplicated hypertension Control: (n = 171), pharmacist intervention: (n = 180), physician intervention: (n = 181), dual intervention: (n = 180). Setting: large, inner‐city, academic, internal medicine practice affiliated with the Indiana University School of Medicine. The primary venues for this study were the general medicine practice and the Wishard Memorial Hospital outpatient pharmacy, which at the time of the study were located 1 floor apart in the Regenstrief Health Centre. Eligibility for this study required that participants had evidence in their electronic medical records of hypertension as an active outpatient diagnosis or, in the absence of such a diagnosis, all of the following: ≥ 2 systolic blood pressure measurements of ≥ 140 mm Hg, ≥ 2 diastolic blood pressure measurements of ≥ 90 mm Hg, and a prescription for ≥ 1 antihypertensive agent. Qualifying antihypertensive agents were ace‐converting enzyme inhibitors, b‐blockers, calcium channel blockers, oral clonidine and topical patch, diuretics, and other less commonly prescribed drugs such as methyldopa and reserpine. Participants were excluded from taking part if they had evidence (diagnoses or test results) indicating the presence of a cardiovascular complication such as coronary artery disease, myocardial infarction, stroke, heart failure, or renal insufficiency. Age mean ± SD: control: 54 ± 11, pharmacist intervention: 54 ± 11, physician intervention: 56 ± 11, dual intervention: 54 ± 11 Gender female n (%): control: 75 (44%), pharmacist intervention: 79 (44%), physician intervention: 78 (43%), dual intervention: 81 (45%) Country: USA Comorbidity: none stated Sociodemographics: potential participants who were able to communicate (could hear and speak English and understand instructions), had access to a working telephone, and were willing to provide written informed consent were enrolled. Formal education, mean ± SD (years), control: 11 ± 3, pharmacist intervention: 10 ± 3, physician intervention: 11 ± 3, dual intervention: 11 ± 3 Married (%), control: 30, pharmacist intervention: 29, physician intervention: 28, dual intervention: 30 Number of people in household, mean ± SD, control: 2.3 ± 1.5, pharmacist intervention: 2.6 ± 1.7, physician intervention: 2.3 ± 1.4, dual intervention: 2.2 ± 1.2 Live alone (%), control: 32, pharmacist intervention: 28, physician intervention: 32, dual intervention: 30. Ethnicity: unknown; control: 57 (33%), pharmacist intervention: 61 (34%), physician intervention: 58 (32%), dual intervention: 58 (32%) | |
| Interventions | Physician intervention The computer‐based ordering system generated care suggestions for both intervention and control groups; however, the suggestions were displayed by the computer to physicians and/or pharmacists for participants randomised to the appropriate intervention groups. This allowed the researchers to assess the numbers and types of interventions that the control group was eligible to receive as well as those in the 3 intervention groups. For participants in the physician intervention group, all care suggestions based solely on earlier Regenstrief medical records system data were generated at the time that the encounter form was printed and were displayed at the end of the drug list. All hypertension care suggestions for intervention participants were displayed as “suggested orders” on physicians’ workstations when they wrote orders after participant visits. This computer screen displayed the actual suggested order, possible actions for each order (order or omit), and a brief explanation of the rationale for the order. Physicians could list full guidelines and literature citations associated with the specific suggestions by using the workstation’s 'help' key. Pharmacist Intervention When any participant brought a new or refill prescription (written in any affiliated clinic, physician’s office, or Wishard Hospital emergency department) to the Wishard outpatient pharmacy, a pharmacy technician entered the data into the Regenstrief medical records system pharmacy module. This was required for all prescriptions because it was the only way to generate and complete a financial transaction for prescriptions in the outpatient pharmacy. After entering prescription data, a high‐speed printer created a label to affix to the participant’s drug container. The technician who filled the prescription notified the pharmacist for all intervention participants. The labelled drug product was checked by a pharmacist who dispensed the agent to the participant and provided counselling. For this study the researchers created the pharmacist intervention recording system. This software programme was used by all Wishard pharmacists to document all pharmaceutical care interventions provided to any outpatient. For participants enrolled in this study only (regardless of study group), care suggestions generated by the Regenstrief medical records system or the outpatient workstations (in response to data entered by the physician, e.g. new antihypertensive prescriptions) were stored in the pharmacist intervention recording system. For participants randomised to receive care from an intervention pharmacist who had such care suggestions, the high‐speed printer printed a note together with drug container labels directing the pharmacist to the pharmacist intervention recording system to display care suggestions that were identical to those viewed by intervention physicians. Physician and pharmacist (dual) Intervention Control group: the control group did not receive any interventions by either physician nor pharmacist | |
| Outcomes | The primary end point was generic health‐related QoL. Secondary end points were symptom profile and side effects from antihypertensive drugs, number of emergency department visits and hospitalisations, blood pressure measurements, participant satisfaction with physicians and pharmacists, drug therapy compliance, and health care charges. | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Participants were taken on a next‐in‐line basis – no random sequence, but sequentially allocated to either intervention or control. Physicians were randomly assigned to practices. There were no details of how randomisation was generated. |
| Allocation concealment (selection bias) | Low risk | Participating physicians and pharmacists were unaware of the study hypothesis. |
| Blinding of participants and personnel (performance bias) | Low risk | Participating physicians and pharmacists were unaware of the study hypothesis. All research assistants and interviewers were blinded to group assignment. |
| Incomplete outcome data (attrition bias) | Unclear risk | All reported |
| Blinding of outcome assessment (detection bias) | Low risk | Data extracted from Regenstrief medical records system records for ED visits, hospitalisations and mortality |
| Selective reporting (reporting bias) | Low risk | All selected outcomes were reported. Random audit (10%) of all paper records from intervention and control groups |
| Protection against contamination bias | Low risk | It is hard to determine if the control group was contaminated with the intervention group. |
| Other bias | Unclear risk | Unclear. There is no evidence of other bias |
| Study characteristics | ||
| Methods | RT Study duration: 13 months | |
| Participants | 701 participants Setting: home Diagnostic criteria: all participants discharged to home following coronary artery bypass graft procedure and/or valve repair or replacement and/or aneurysm repair, or other cardiac procedure Age (years) (mean ± SD): intervention group: 62.8 (10.6), control group: 63.2 (10.9) Sex female n (%): intervention group: 73 (21.5%), control group: 88 (24.4%) Country: USA Comorbidity n (%): diabetes mellitus: intervention 123 (34.0), control 111 (32.6); hypertension: intervention 268 (74.2), control 283 (83.2); dyslipidaemia: intervention 263 (72.8), control 274 (80.5); dialysis: intervention 8 (2.2), control 7 (2.0); cerebrovascular accident: intervention 15 (4.1), control 9 (2.6); COPD: intervention 44 (12.1), control 30 (8.8); peripheral vascular disease: intervention 29 (8.0); control 25 (7.3); previous myocardial infarction: intervention 146 (40.4), control 144 (42.3); CHF: intervention 51 (14.1), control 48 (14.1); arrhythmia: intervention 37 (10.2), control 39 (11.4) Sociodemographics: not stated Ethnicity: intervention group: 289 (84.4%) white, 53 (15.5%) non‐white; control group: 318 (88%) white, 43 (11.9%) non‐white Date of study: August 2009‐September 2011 | |
| Interventions | 1 intervention group 340 participants, control 361 participants Hospital‐employed, cardiothoracic physician assistants conducted home visits on post‐discharge days 2 and 5, with occasional variation due to participant availability and Sundays, on which no house calls were made. The same hospital‐based physician assistants responsible for perioperative and intraoperative care were assigned to make house calls. During a house call, the physician assistant performed a focused physical exam and reviewed the participant’s medications. Adjustments were made to the participant’s medications, and new medications were prescribed as necessary. The surgical wounds were examined carefully and all participant concerns were addressed. Prescriptions were written for antibiotics, blood work, or imaging studies when indicated. Arrangements were made if the participant needed to be evaluated as an inpatient. All findings were documented on the visit form. Both groups were seen in the office on post‐discharge weeks 2 and 4. The control group was seen at home by standard visiting nurses without any specialty training or expertise in caring for people with cardiac surgery | |
| Outcomes | Hospital admissions/number of people admitted to hospital | |
| Notes | Not sure of randomisation and these were hospital‐based physician assistants working in homes. | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Unclear from the document |
| Allocation concealment (selection bias) | Unclear risk | Unclear from the document |
| Blinding of participants and personnel (performance bias) | Unclear risk | Unclear from the document |
| Incomplete outcome data (attrition bias) | High risk | 19% of the participants in the intervention group refused to participate or failed to respond to requests to participate. |
| Blinding of outcome assessment (detection bias) | Low risk | This is less important for our purpose as we are looking at objective outcomes (hospitalisations, ED visits, and mortality). |
| Selective reporting (reporting bias) | Low risk | No evidence of this |
| Protection against contamination bias | Low risk | No contamination |
| Other bias | Unclear risk | Unclear. There is no evidence of other bias |
| Study characteristics | ||
| Methods | Prospective, randomised, comparative study Duration: 6 months | |
| Participants | 330 participants with mild‐to‐moderate essential hypertension Age, years (mean ± SD): intervention 61.95 ± 11.4, control 61.71 ± 11.3, P = 0.85 Sex female n (%): intervention 72 (44%), control 90 (54%) Country: USA, California No statistically significant differences were noted between the groups. Concurrent disease (number of participants) intervention 98, control 95, P = 0.74 Smoker (number of participants): intervention 15, control 9, P = 0.18 Alcohol consumer (number of participants): intervention 12, control 9, P = 0.47 Sociodemographics: not reported Ethnicity: not reported | |
| Interventions | Hypertension care provided by either the pharmacist‐managed hypertension clinic or physician‐managed general medical clinics In the pharmacist‐managed hypertension clinic, a clinical pharmacist managed the treatment of participants, who made up the experimental group. Physicians were contacted and provided consent for any therapeutic changes but were asked not to adjust drug therapy unless a lack of intervention would be dangerous for the participant. In the physician‐managed clinic, physicians managed the treatment of participants independently with no pharmacy intervention; this was the control group. Participants randomly assigned to the pharmacist‐managed hypertension clinic group were counselled by the clinical pharmacist. The pharmacist informed the participants that an effort would be made to decrease the number of drugs they took for hypertension or to alter their therapy by administering more appropriate or less expensive drugs to achieve similar or improved blood pressure control. The pharmacist determined the most appropriate antihypertensive regimen for the participant and ordered laboratory tests as needed. The pharmacist also provided education on nonpharmacologic ways to control blood pressure. Control group: participants randomised to the physician‐managed clinic group were referred back to their primary care provider for hypertension treatment. These participants received no intervention, and physicians treated them in the customary manner | |
| Outcomes | Number of hospitalisations | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Randomisation, but no description of sequence generation |
| Allocation concealment (selection bias) | Unclear risk | Not clear. The document does not state whether the allocation was concealed or not |
| Blinding of participants and personnel (performance bias) | Unclear risk | It is unclear if participants were blinded to the intervention. |
| Incomplete outcome data (attrition bias) | Low risk | Attrition rates in both groups are comparable for various reasons. |
| Blinding of outcome assessment (detection bias) | Unclear risk | Method of data collection is not clear. |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported as mentioned at the start of the trial. |
| Protection against contamination bias | Unclear risk | It is hard to determine if the intervention group was contaminated with the control group. |
| Other bias | Unclear risk | It would have been more desirable to have only newly diagnosed participants, but the sample size was already small. Results cannot be extrapolated to other physician groups. |
| Study characteristics | ||
| Methods | RT Duration: not provided | |
| Participants | 630 participants, 315 intervention, 315 control. Final sample analysed 253 intervention, 264 control Setting: pharmaceutical care was provided at home Diagnostic criteria: participants aged ≥ 65 on 5 prescription medications taken without assistance | |
| Interventions | Organisational Participants in the ‘pharmaceutical care’ group were visited at home by a pharmacist at the beginning of the project. The pharmacist examined the medicines list with regard to possible side‐effects, interactions, and administration, then tried to make the regime less complex, informed the participants meanwhile about the drugs, listened to questions concerning the drugs, handed over information leaflets, and motivated adherence. Nine different pharmacists were involved and adhered to the Danish manual for pharmaceutical care: ‘Medication Review ‐ Managing Medicine Manual’. The aim of the ‘Medication Review ‐ Managing Medicine’ is to prevent, identify, and resolve drug‐related problems and to contribute to rational pharmacotherapy for participants and society. Control participants were not provided any intervention. | |
| Outcomes | The primary endpoint was treatment adherence assessed by a pill‐count in all participants during 1 year. Only oral prescription drugs taken throughout the study period were included in the adherence calculation. In addition, a project nurse visited all participants initially, then at 6 and 12 months to photograph pills to be counted later by a ‘counter pen’ (a combination of a marker and a digital counter). The adherence rate (%) per drug was calculated as mean adherence rate during 1 year. We also calculated adherence rates for the intervals of 0‐6 and 6‐12 months. Secondary outcome measures included drug‐related problems, hospitalisations and mortality measured during the intervention year and at 2‐year follow‐up. | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Participants were randomly assigned to control and intervention groups. |
| Allocation concealment (selection bias) | Unclear risk | No mention is made of allocation concealment. |
| Blinding of participants and personnel (performance bias) | High risk | Pharmacist was aware of whether the participant was in the intervention group. |
| Incomplete outcome data (attrition bias) | Low risk | Little differential attrition between intervention and control groups. |
| Blinding of outcome assessment (detection bias) | Low risk | Hospitalisations and mortality are objective outcomes. |
| Selective reporting (reporting bias) | Low risk | No evidence of selective reporting |
| Protection against contamination bias | Low risk | Control group participants were not provided any pharmaceutical intervention. |
| Other bias | Unclear risk | According to the study authors, control participants were not exposed to any intervention, but something was done and this was not specified. |
| Study characteristics | ||
| Methods | Prospective, randomised, controlled, longitudinal, 2‐year pilot study | |
| Participants | 104 participants, nonprofit university‐affiliated dialysis clinic Participants > 18 years with end‐stage renal disease who were undergoing a stable haemodialysis regimen for at least 3 months Age (yrs ± SD): intervention 56.3 ± 15, control 60.5 ± 14.7 Sex female n (%): intervention 22 (39), control 28 (60), P < 0.03 Country: USA, New Mexico Baseline clinical characteristics: length of time participant had been receiving haemodialysis (years): intervention 2.8 ± 1.8, control 2.4 ± 2.2 Number of drugs used: intervention 10 ± 4, control 10 ± 4 Cost of drugs (USD); intervention 430 ± 197, control 451 ± 267 Comorbidity: end‐stage renal disease aetiology Diabetes mellitus n (%): intervention 22 (39), control 23 (49) Hypertension n (%): intervention 18 (32), control 12 (26) White n (%): intervention 13 (23), control 16 (34) Hispanic n (%): intervention 17 (30), control 15 (32) Native American n (%): intervention 13 (23), control 5 (11) | |
| Interventions | Intervention group: effects of pharmaceutical care, consisting of 1‐1, in‐depth drug therapy reviews conducted by a clinical pharmacist, versus Control group: standard care, consisting of brief drug therapy reviews conducted by a nurse on several participant outcomes in ambulatory participants undergoing haemodialysis. Participants assigned to pharmaceutical care had drug therapy reviews conducted by a nephrology‐trained clinical pharmacist or 1 of 2 pharmacists completing postdoctoral training in nephrology pharmacotherapy. Types of drug‐related problems were recorded and evaluated by using a previously described method. All drug‐related problems were assigned to 10 possible categories: untreated indications, improper drug selection, sub therapeutic dosage, overdose, adverse drug reactions, drug interactions, failure to receive drugs, medical record discrepancy, inadequate education of participant or health care professional, and drug use without indication. The drug‐related problems were further categorised into therapeutic drug classes, and the outcome related to the drug‐related problem intervention was captured. The standard care group served as the control group. The participants in the standard care group received periodic drug profile updates by dialysis nursing staff as mandated by the dialysis clinic policy and procedure. These are typically brief interactions in which participants are queried as to whether any drugs have changed since the last review. | |
| Outcomes | Mortality | |
| Notes | The study experienced high attrition due to death, transplantation, or transfer to a different facility, with about 50% of participants remaining at the end of study. The study also did not conduct an assessment of the relationship between drug‐related problem resolution and hospitalisations, which could provide useful information as to whether targeted pharmaceutical care interventions would be helpful. | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Randomly assigned, by dialysis shift but no description of sequence generation |
| Allocation concealment (selection bias) | Low risk | Randomisation was conducted by the clinic nurse manager, who had no affiliation with the study, by drawing the shift name from an opaque envelope and assigning the first 3 drawn shifts to pharmaceutical care and the remainder to standard care. |
| Blinding of participants and personnel (performance bias) | Low risk | Blinding as far as participants and personnel were not communicating ‐ different shifts |
| Incomplete outcome data (attrition bias) | High risk | There was a high level of attrition due to death, transplantation or transfer to a different facility, with about 50% of participants remaining at the end of the study. |
| Blinding of outcome assessment (detection bias) | Low risk | Outcome data on hospital admissions were provided by hospital episode statistics (not self‐report) and are therefore unlikely to be biased. |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported as mentioned at the start of the trial. |
| Protection against contamination bias | Low risk | It can not be determined if the intervention group was contaminated with the control group. |
| Other bias | Unclear risk | Unclear. It is unclear if there is other bias. |
| Study characteristics | ||
| Methods | Cluster‐RT Study duration: 12 months | |
| Participants | 3230 participants Setting: nursing homes Diagnostic criteria: none provided Age: participant characteristics not provided in terms of mean age, just percent of sample in intervention and control groups that were in particular age ranges Sex: participant characteristics not provided Country: Australia Comorbidity: not provided Ethnicity: participant characteristics not provided Date of study: unknown although paper was initially received for publication in May 2000 | |
| Interventions | 1 intervention group The 12‐month intervention involved 3 phases: introducing a new professional role to stakeholders with relationship building, nurse education, and medication review by pharmacists who had a postgraduate diploma in clinical pharmacy. The clinical pharmacy service model introduced to each nursing home was supported with activities such as focus groups facilitated by a research nurse, written and telephone communication, and face‐to‐face professional contact between nursing home staff and clinical pharmacists on issues such as drug policy and specific resident problems, together with education and medication review. This was a multifaceted intervention directly targeting nursing homes. Most of the contact with GPs was indirect using the existing relationships between nursing homes and visiting GPs. A number of focus groups and personal interviews about the project were conducted with GPs. Control nursing homes continued with usual care. | |
| Outcomes | Mortality was collected at the end of the 12‐month study. | |
| Notes | No significant changes were observed in annual mortality rates or frequency of hospitalisations between intervention and control nursing home groups. It is unclear from Table 5, which shows the mortality and hospitalisation data, how the study authors arrived at their figures or their conclusions. Therefore, we were unable to use the data to calculate hospitalisations. | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Nursing homes were selected for the intervention treatment by random draws from a hat. |
| Allocation concealment (selection bias) | Low risk | Not clear if this was done although the homes were independently assigned to the control or intervention groups. However, according to the EPOC criteria, the risk of bias for this study is low because units, in this case nursing homes, were assigned rather than individuals. |
| Blinding of participants and personnel (performance bias) | Unclear risk | Unclear, although with the objective outcomes that we are interested in this is less of a concern (according to EPOC risk of bias criteria). |
| Incomplete outcome data (attrition bias) | Low risk | Control and intervention groups did not appear to differ in terms of attrition. |
| Blinding of outcome assessment (detection bias) | Low risk | This is less important for our purpose as we are looking at objective outcomes (hospitalisations, ED visits, and mortality). |
| Selective reporting (reporting bias) | Low risk | No evidence of this |
| Protection against contamination bias | Low risk | There is indication to suggest that the intervention was contaminated by the control group. |
| Other bias | Unclear risk | The limited duration of the study and size of the sample may have compromised the ability to detect an effect. We considered other bias due to cluster randomisation. |
| Study characteristics | ||
| Methods | RT Study duration: 26 weeks | |
| Participants | 148 intervention, 145 control Setting: primary care Age: median, intervention 84 years, control 83 years Sex female n (%): intervention 66%, control 66% Country: Denmark Diagnosis: cardiovascular disease: intervention 45 (30%), control 28 (19%); other intervention 103 (70%), control 117 (81%); P = 0.02 Sociodemographics: housing: living in private home intervention: 95%, control 97%; widow/widower: intervention 59%, control 57%; married: intervention 30%, control 29%; divorced/single: intervention 11%, control 14% Ethnicity: unclear, possibly white Danish Date of study: November 2003‐June 2005 | |
| Interventions | 1 intervention and 1 control The intervention follow‐up consisted of 3 contacts. The main intervention was a joint home visit involving both the GP and the district nurse. It was conducted approximately 1 week after discharge and was guided by an agenda. Control group: standard care | |
| Outcomes | The primary outcome measures were hospital readmissions of any kind and the concordance between the GP's knowledge of the medical treatment and what the participant was actually taking. | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Randomisation was generated by a computer. |
| Allocation concealment (selection bias) | Low risk | Computer |
| Blinding of participants and personnel (performance bias) | Unclear risk | This would be impossible |
| Incomplete outcome data (attrition bias) | Low risk | All participants were reported |
| Blinding of outcome assessment (detection bias) | Low risk | Independent team |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported |
| Protection against contamination bias | Low risk | There is no indication to suggest that the intervention was contaminated by the control group. |
| Other bias | Unclear risk | It is unclear if there are other risks not accounted for. |
| Study characteristics | ||
| Methods | RT Study duration: 201 days (3‐week intervention and 180 days of follow‐up) | |
| Participants | 154 participants Setting: home Diagnostic criteria: participants had to have a primary or secondary diagnosis of heart failure and were referred to receive skilled nursing services Age (years) (mean ± SD): control: 78.1 (11.2), intervention: 81.3 (9.3), participants had to be ≥ 21 years Female n (%): control: 55 (72), intervention: 56 (73) Country: USA Comorbidity: heart failure Sociodemographics: the catchment area provided participants from urban, suburban, and rural environs and from across all socioeconomic classes. According to census data, 89% of the population of these 3 counties combined is white, and 87% of adults have a high school diploma. Median household income for the counties is approximately USD 47,000 (2003 data). Non‐English‐speaking participants were included if adequate translation services were available from family members or friends. Ethnicity: unknown; it is difficult to ascertain the ethnicity from the information given. control 68 (88%), intervention 75 (97%) Date of study: 1 July 2002 to end 2004 | |
| Interventions | 1 intervention: pharmaceutical care services Pharmaceutical care services consisted of an initial comprehensive in‐home medication assessment (concurrent with agency admission) and 2 follow‐up visits (7‐10 and 18‐21 days later). The follow‐up visits were contingent on the participant’s continued receipt of visiting nurse services (i.e. participants discharged from the visiting nurse before 21 days would not receive all of the pharmacist’s planned visits). Throughout the 3‐week intervention period, the clinical pharmacist accessed and reviewed all pertinent physician notes and laboratory test values via the National Endowment for the Humanities data system and interacted with prescribers on behalf of the participants as necessary. Control participants received the usual care provided by the visiting nurse association. Visiting nurse services (provided to both groups) included basic nursing care and a brief physical assessment and medical history. | |
| Outcomes | Hospitalisations and mortality were assessed during a 180 day follow‐up period. | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | This was adequate: "Patients provided informed consent for study participation and were randomised to receive usual care or usual care plus pharmaceutical care by means of a computer‐generated random numbers table in blocks of four." |
| Allocation concealment (selection bias) | Low risk | This was adequate: "Once informed consent was received, the nurse obtained a baseline quality‐of life assessment (using the SF‐12) and then accessed a sealed envelope containing the group assignment from the intake office." |
| Blinding of participants and personnel (performance bias) | Low risk | Blinding or a lack of blinding was unlikely to affect the outcome because the usual care group received usual care from nurses whereas the intervention group received usual care plus the services of a pharmacist |
| Incomplete outcome data (attrition bias) | Low risk | Attrition in both the intervention and control groups was comparable |
| Blinding of outcome assessment (detection bias) | Low risk | This is of less concern in our case as all of our outcomes are objective (according to the EPOC risk of bias criteria). |
| Selective reporting (reporting bias) | Low risk | No evidence of a problem |
| Protection against contamination bias | Low risk | There is no evidence of contamination between the intervention and the control groups |
| Other bias | High risk | Small sample may have produced low power to detect an effect. Poor pharmacist‐prescriber communication may have reduced efficacy of intervention. |
| Study characteristics | ||
| Methods | RT of clinical medication review by a pharmacist against normal general practice review Length of study: 12 months (study conducted: June 1999‐June 2000) | |
| Participants | Participants from general practices 1188 participants aged ≥ 65 or over who were receiving at least 1 repeat prescription and living in the community. Age: mean (SD) intervention, 74 (6.6) control, 73 (6.4) Sex female n (%): intervention 339 (56%), control 325 (56%) Country: UK Comorbidity: not reported Sociodemographics: not reported Ethnicity: not reported | |
| Interventions | 1 intervention and 1 control; 601 participants in the intervention and 580 participants in the control group. Intervention group: participants were invited to a consultation at which the pharmacist reviewed their medical conditions and current treatment according to a specific algorithm which includes history taking and data gathering, evaluation and implementation stages. Control group: participants in the control group continued to receive normal care from their GP and primary healthcare staff. Participants were recalled for review of treatment by the GP according to normal custom in the practice. | |
| Outcomes | Hospital admission and mortality | |
| Notes | ||
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Randomisation was done by a computer. |
| Allocation concealment (selection bias) | Low risk | Practice based allocation. |
| Blinding of participants and personnel (performance bias) | Unclear risk | This would be impossible. |
| Incomplete outcome data (attrition bias) | Low risk | Some participants did not complete the study and were reported in the study. |
| Blinding of outcome assessment (detection bias) | Unclear risk | No explanation given. |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported. |
| Protection against contamination bias | Unclear risk | It is unclear if there was a contamination between the intervention and the control group. |
| Other bias | Unclear risk | It is unclear to determine if there are other biases. |
| Study characteristics | ||
| Methods | An open randomised controlled trial of clinical medication review by a pharmacist of elderly care home residents against usual care Study duration: 6 months | |
| Participants | 661 participants (331 intervention group but only 315 received Intervention) and 330 participants in the control group Setting: aged care facilities in Leeds, UK (nursing, residential and mixed care homes for older people in Leeds, UK) Diagnostic criteria: residents aged ≥ 65, seeking to recruit all residents taking ≥ 1 repeat medicines Age (years) (mean ± SD): age mean (interquartile range), Intervention 85.3 (81 to 90) and control 84.9 (80 to 90) Sex male n (%): intervention 75 (22.7), control 79 (23.9) Country: UK Comorbidity: not stated Sociodemographics: not stated Ethnicity: not stated Date of study: not reported, but paper first published 12/8/2006 | |
| Interventions | 1 intervention A clinical medication review was conducted by the study pharmacist within 28 days of randomisation. It comprised a review of the general practice clinical record and a consultation with the participant and carer. The pharmacist formulated recommendations with the participant and carer and passed them on a written proforma to the GP for acceptance and implementation. GP acceptance was signified by ticking a box on the proforma. Control participants received usual GP care | |
| Outcomes | The primary outcome measure was the number of changes in medication per participant. Secondary outcome measures were the following:
| |
| Notes | Randomisation was curtailed on 30 June 2003 when it became clear that the intended sample size was not achievable within the available timescale. Data were analysed on an intention‐to‐treat basis. | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Participants were randomly sized blocks. |
| Allocation concealment (selection bias) | Unclear risk | This is not mentioned in the study. |
| Blinding of participants and personnel (performance bias) | Unclear risk | It is not clear if the participants were blinded to the intervention. |
| Incomplete outcome data (attrition bias) | Low risk | Some participants did not complete the study and were reported in the study. |
| Blinding of outcome assessment (detection bias) | Low risk | A nurse blind to the study assessed participants. |
| Selective reporting (reporting bias) | Low risk | All outcomes were reported. |
| Protection against contamination bias | Unclear risk | It is difficult to determine if the intervention group was contaminated by the control group. |
| Other bias | Unclear risk | It is difficult to determine if there are other biases. |
ADL: activities of daily living; BPMDL: best possible medication discharge list; CI: confidence interval; CHAP: Cardiovascular Health Awareness Program; CHD: coronary heart disease; CHF: congestive heart failure; COPD: chronic obstructive pulmonary disease; ED: emergency department; EuroQoL‐5D: EuroQol Group Association ("The EuroQol Group") comprises a network of international, multilingual, multidisciplinary researchers; EQ‐5D: a standardised instrument for use as a measure of health outcome; GP: general practitioner; ICU: intensive care unit; QoL: quality of life; RT: randomised trial; SD: standard deviation; SF‐12: Short Form‐12; SF‐36: Short Form‐36; STOPP: Screening Tool of Older Persons Prescriptions; START: Screening Tool to Alert Doctors to Right Treatment
Characteristics of excluded studies [ordered by study ID]
Ir a:
| Study | Reason for exclusion |
| Irrelevant intervention | |
| Not primary care | |
| Protocol of a study | |
| This study does not appear to be a primary care intervention. | |
| Outcomes not relevant | |
| Outcomes not relevant | |
| Outcomes not relevant | |
| Not primary care | |
| Irrelevant intervention | |
| Study protocol | |
| Outcomes not relevant | |
| This was not a primary care intervention; it was done by geriatricians. | |
| Not primary care | |
| Irrelevant intervention | |
| Outcomes not relevant | |
| Outcomes not relevant | |
| Outcomes not relevant | |
| Cost‐effectiveness study that had data in a form not enabling data extraction. | |
| Study protocol | |
| Outcomes not relevant | |
| Study protocol | |
| Study protocol | |
| Irrelevant intervention | |
| Study was for a pre‐post design not included in the protocol | |
| Outcomes not relevant | |
| This study did not occur in primary care. | |
| Study was for a pre‐post design not included in the protocol | |
| Outcomes not relevant | |
| Outcomes not relevant | |
| This study did not occur in primary care. | |
| This is not a randomised trial (it is described as a controlled intervention study and there is no evidence of randomisation). | |
| Outcomes not relevant | |
| Not primary care | |
| Not possible to extract appropriate data | |
| This study was not a randomised trial. | |
| This study was not a randomised trial. | |
| This study was not a randomised trial. | |
| This study was a conference abstract only and did not address adverse drug reactions. | |
| Outcomes not relevant | |
| This study is reported elsewhere (see Furniss 2000, also excluded) | |
| Not a randomised trial | |
| Not a randomised trial | |
| This is a hospital intervention and not done in primary care. It appears that the study pharmacist is recruited from the hospital. As it says, the pharmacist complied with the Society of Hospital Pharmacist clinical pharmacy services. It is also published in a hospital journal. | |
| Not primary care | |
| Not a randomised trial | |
| Outcomes not relevant | |
| Study protocol | |
| Irrelevant intervention | |
| Study protocol | |
| This study was not a randomised trial | |
| This study was not a randomised trial | |
| The outcomes reported are not appropriate for this study | |
| Outcomes not relevant | |
| Study protocol | |
| Reported elsewhere (see Bernsten 2001) | |
| Outcomes not relevant | |
| Study protocol | |
| Not primary care | |
| Complex study, which made data extraction not possible. |
Data and analyses
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Number of hospital admissions Show forest plot | 2 | 3889 | Risk Ratio (M‐H, Random, 95% CI) | 1.24 [0.79, 1.96] |
| Analysis 1.1  Comparison 1: Professional interventions versus standard care, Outcome 1: Number of hospital admissions | ||||
| 1.2 Number of people admitted to hospital Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| Analysis 1.2  Comparison 1: Professional interventions versus standard care, Outcome 2: Number of people admitted to hospital | ||||
| 1.3 Number of emergency department visits Show forest plot | 2 | 1067 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.71 [0.50, 1.02] |
| Analysis 1.3  Comparison 1: Professional interventions versus standard care, Outcome 3: Number of emergency department visits | ||||
| 1.4 Mortality Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| Analysis 1.4  Comparison 1: Professional interventions versus standard care, Outcome 4: Mortality | ||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 2.1 Number of hospital admissions Show forest plot | 11 | 6203 | Risk Ratio (M‐H, Random, 95% CI) | 0.85 [0.71, 1.03] |
| Analysis 2.1  Comparison 2: Organisational interventions versus standard care, Outcome 1: Number of hospital admissions | ||||
| 2.2 Number of people admitted to hospital Show forest plot | 13 | 152237 | Risk Ratio (M‐H, Random, 95% CI) | 0.92 [0.86, 0.99] |
| Analysis 2.2  Comparison 2: Organisational interventions versus standard care, Outcome 2: Number of people admitted to hospital | ||||
| 2.3 Number of emergency department visits Show forest plot | 5 | 1819 | Risk Ratio (M‐H, Random, 95% CI) | 0.75 [0.49, 1.15] |
| Analysis 2.3  Comparison 2: Organisational interventions versus standard care, Outcome 3: Number of emergency department visits | ||||
| 2.4 Mortality Show forest plot | 12 | 154962 | Risk Ratio (M‐H, Random, 95% CI) | 0.94 [0.85, 1.03] |
| Analysis 2.4  Comparison 2: Organisational interventions versus standard care, Outcome 4: Mortality | ||||
Study flow diagram
Risk of bias summary: review authors' judgements about each risk of bias item for each included study
Funnel plot of comparison: 2 Organisational interventions, outcome: 2.1 Number of hospital admissions
Funnel plot of comparison: 2 Organisational interventions, outcome: 2.2 Number of people admitted to hospital
Funnel plot of comparison: 2 Organisational interventions, outcome: 2.3 Number of emergency department visits
Funnel plot of comparison: 2 Organisational interventions, outcome: 2.4 Mortality
Comparison 1: Professional interventions versus standard care, Outcome 1: Number of hospital admissions
Comparison 1: Professional interventions versus standard care, Outcome 2: Number of people admitted to hospital
Comparison 1: Professional interventions versus standard care, Outcome 3: Number of emergency department visits
Comparison 1: Professional interventions versus standard care, Outcome 4: Mortality
Comparison 2: Organisational interventions versus standard care, Outcome 1: Number of hospital admissions
Comparison 2: Organisational interventions versus standard care, Outcome 2: Number of people admitted to hospital
Comparison 2: Organisational interventions versus standard care, Outcome 3: Number of emergency department visits
Comparison 2: Organisational interventions versus standard care, Outcome 4: Mortality
| Professional interventions compared to standard/usual care for prevention of medication errors | ||||||
| Patient or population: adults receiving medication in primary care | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with standard/usual care | Risk with professional interventions | |||||
| Number of hospital admissions | Study population | RR 1.24 | 3889 | ⊕⊕⊕⊝ | The two studies had wide confidence intervals. | |
| 17 per 1000 | 21 per 1000 | |||||
| Number of people admitted to hospital | Study population | RR 0.99 | 3661 | ⊕⊕⊕⊕ | ||
| 448 per 1000 | 443 per 1000 | |||||
| Number of emergency department visits | Study population | RR 0.71 | 1067 | ⊕⊕⊝⊝ | The two studies had wide confidence intervals and selection bias. | |
| 118 per 1000 | 85 per 1000 | |||||
| Mortality | Study population | RR 0.98 | 3538 | ⊕⊕⊕⊝ | ||
| 122 per 1000 | 119 per 1000 | |||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1We downgraded one level due to imprecision. | ||||||
| Organisational interventions compared to standard/usual care for prevention of medication errors | ||||||
| Patient or population: adults receiving medication in primary care | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Certainty of the evidence | Comments | |
|---|---|---|---|---|---|---|
| Risk with standard/usual care | Risk with organisational interventions | |||||
| Number of hospital admissions | Study population | RR 0.85 | 6203 | ⊕⊝⊝⊝ | Some studies had unclear risk of bias (selection and attrition), high heterogeneity and wide confidence intervals. | |
| 274 per 1000 | 233 per 1000 | |||||
| Number of people admitted to hospital | Study population | RR 0.92 | 152,237 | ⊕⊕⊝⊝ | Some studies had unclear risk of bias (selection, attrition and performance bias) and wide confidence intervals. | |
| 13 per 1000 | 13 per 1000 | |||||
| Number of emergency department visits | Study population | RR 0.75 | 1819 | ⊕⊝⊝⊝ | Studies had unclear risk of bias (selection, performance and attrition bias), high heterogeneity and wide confidence intervals. | |
| 234 per 1000 | 176 per 1000 | |||||
| Mortality | Study population | RR 0.94 | 154,962 | ⊕⊝⊝⊝ | Studies had high risk of selection, attrition and performance bias and wide confidence intervals. | |
| 50 per 1000 | 47 per 1000 | |||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1We downgraded one level for unclear risk of bias (selection and attrition bias). | ||||||
| Study | Name | Theory | Materials | Procedures | Who provided intervention | Modes of delivery |
|---|---|---|---|---|---|---|
| Pharmaceutical care | Pharmaceutical care is the provision of drug therapy for the purpose of achieving outcomes that improve a person’s quality of life. | Pharmacies in the intervention group provided pharmaceutical care, which consisted of offering the pharmaceutical care service to participants and to their corresponding GPs. | An Initial interview and assessment of the therapeutic plan was undertaken, registration of data during the subsequent visits to allow the identification of medication‐related problems, and an intervention to solve the problem. The intervention involved proposing changes in the medication participants received, which had to be communicated to the patient’s GP. | Pharmacists provided the intervention. | Individual and face‐to‐face | |
| Pharmaceutical care | Pharmaceutical care is the provision of drug therapy for the purpose of achieving outcomes that improve a person's quality of life, although little research has been conducted in community‐based pharmaceutical care with elderly people. | Training of pharmacists was done with a study manual. The manual contained an overview of the concept of pharmaceutical care and its provision to elderly people. No reference was provided for the study manual. | The intervention group of pharmacists identified actual and potential drug‐related problems using a structured approach. These pharmacists utilised a number of data sources in this assessment including the participant, the participant's GP, and pharmacy records. Following this assessment, pharmacists were instructed to formulate an intervention and monitoring plan. | Community pharmacists were trained to provide the structured pharmaceutical care intervention. A study manual helped facilitate this process. It contained an overview of the concept of pharmaceutical care, its provision to elderly people, information on the therapeutic management of a number of disease states common in the elderly, together with other issues pertinent to drug therapy in the elderly. | Individual face‐to‐face | |
| Drug evaluation and recommendation | Several instruments, criteria, and algorithms have been developed to enable more rational and appropriate use of medication, but limited evidence exists with regard to the outcomes that were investigated. | The Good Palliative‐Geriatric Practice algorithm Garfinkel 2007) and the STOPP/START criteria were used (O'Mahony 2015). Both of these tools assess the appropriate use of medication in older people. | The intervention was composed of 3 phases. In the first phase, an experienced pharmacist evaluated all prescriptions using the GP‐GP algorithm and based their decision about appropriateness on the STOPP/START criteria. In the second phase, the pharmacist discussed recommendations for each drug with the participant's physician in order to come up with a final list of recommendations. Finally, the recommendations were discussed with the participant and a final decision was agreed by physicians and participants. | The intervention was delivered by a trained and experienced pharmacist. No details are provided concerning what is a "trained and experienced" pharmacist. | Individual and face‐to‐face | |
| Chronic care clinics | Chronic care clinics redesign the structure and content of primary care services through the delivery of scheduled visits devoted to chronic disease management. This mode of service delivery has the potential to improve outcomes for elderly people. | The chronic care clinics included an extended visit with the physician and nurse dedicated to planning chronic disease management, a pharmacist visit that emphasised reduction of polypharmacy and high‐risk medications, and a patient self‐management group. | Frail older people were invited to participate in visits with the primary care team. During these visits, a shared treatment plan was developed, a session was conducted with the pharmacist that addressed polypharmacy and medications associated with functional decline, patient self‐management group sessions were conducted, and the provision of health status assessment information was provided to the practice team. | The team that provided the intervention consisted of the participant’s physician, a team nurse, and a pharmacist. Physicians and team nurses received training in population‐based medicine and management strategies of geriatric syndromes. Team nurses received on‐the‐job coaching from study staff. | The intervention was delivered individually and in groups in a face‐to‐face format. | |
| Medication review and drug recommendations | Potentially inappropriate prescriptions are prevalent in older people and are associated with adverse drug events. The STOPP/START criteria are designed to detect potentially inappropriate prescriptions in elderly people. However, little is known about the effects of an intervention involving the application of the STOPP/START criteria on clinical outcomes. | The STOPP/START criteria were used to deliver the intervention (Gallagher 2008). The STOPP criteria focus on avoiding the use of drugs that are potentially inappropriate for older people and the START criteria identify undertreatment or prescribing omissions in older people. | Medication reviews were conducted by the study pharmacist for all residents. Recommendations made by the pharmacist were discussed with the chief physician. The physician then decided whether to accept these recommendations and implement prescribing changes. | The intervention was conducted by the study pharmacist who applied the STOPP/START criteria during the medication review. The pharmacist also discussed the recommendations from the intervention with the chief physician, who decided whether to accept these recommendations and implement prescribing changes. | Intervention was delivered individually and face‐to‐face. | |
| Structured educational intervention | Inappropriate drug prescription is a common problem in people living in nursing homes and is linked to adverse health outcomes. This study assessed the effect of an educational intervention directed to nursing home physicians in reducing inappropriate prescription and improving health outcomes and resource utilisation. | Educational material and references were given to physicians and two 1‐h workshops were used to review cases and promote practice changes. The STOPP/START criteria were reviewed with a random sample of 10 residents cared for by each physician (Gallagher 2008). The content of the educational intervention is provided in an appendix (Garcia‐Gollarte 2014). | The educational intervention included general aspects of prescription and drug use in geriatric patients, how to reduce the number of drugs, to perform a regular review of medications, to avoid inappropriate drug use, to discontinue drugs that do not show benefits, and to avoid under treatment with drugs that have shown benefits. It also discussed some drugs frequently related to adverse drug reactions in older people. | A nursing home physician delivered the structured educational intervention. | Face‐to‐face intervention delivered in a group and individual format. | |
| Medicine reconciliation and action plan | Emergency department overcrowding has been linked to increased mortality, costs, and length of stay. This study evaluated the effectiveness of a telephone‐based, medicines‐management service on reducing emergency department utilisation. | Medication therapy management was provided to participants (APA 2008). A pharmacy technician completed telephonic medication reconciliation, after which a trained pharmacist consulted with the participant or caregiver via telephone to complete a scheduled, comprehensive medication therapy review to identify and resolve any medication‐related problems. The pharmacist constructed a personal medication record and a medication‐related action plan for the participant. The action plan was a participant‐centred document that assisted participants, caregivers, and the pharmacist in the resolution of identified medication‐related problems. | The intervention commenced with a pharmacy technician completing medication reconciliation with the participant over the telephone. Then, a pharmacist consulted with the participant by telephone for an average of 30 min to complete a comprehensive medication review to identify and resolve medication‐related problems. The pharmacist constructed a person medication‐related action plan and followed‐up with the participant's prescriber. | A pharmacy technician delivered the initial medicine reconciliation with the participant. A trained pharmacist conducted the medication therapy review, constructed a personal medication record, and a medication‐related action plan. The pharmacist also followed up with the participant's prescriber for resolution of problems that could not be resolved with the participant. | The intervention was conducted individually on the telephone. | |
| Automated system to facilitate flow of information and provide warnings, alerts, and recommendations | Transitions between the impatient and outpatient setting is a period of high risk for older adults. Most approaches to improving transitions require a substantial commitment of resources but automating these processes may improve the quality and safety of care. | An automated system was used to facilitate the flow of information to the medical group's primary care providers about individuals who were discharged to home from the hospital (Field 2012). | An automated system was developed to facilitate the flow of information to the medical group's primary care providers. A computer interface linked the primary care provider's electronic health records to the hospital records, which provided information about admissions and discharges. The system also provided information about new drugs at discharge, warnings about drug‐drug interactions, recommendations about dose changes and laboratory monitoring of high‐risk medications, and alerts to the provider's support staff to schedule a post‐hospitalisation office visit within 1 week of discharge if not already scheduled. | The automated system delivered the intervention. | The intervention was delivered electronically. | |
| Care transitions clinic visit | Medication errors related to hospital discharge result in rehospitalisations and emergency department visits, which may be reduced by pharmacist involvement during postdischarge transitions of care. This study evaluated the impact of a transitional care clinic visit conducted by a pharmacist. | The Best Possible Medication Discharge List was used to identify medication discrepancies (Wong 2008). It served as the gold standard for the list of medications that the participant should take after discharge. | Participants in the intervention group were scheduled for a care transitions clinic visit approximately 72 h after hospital discharge. The visit involved performing a complete medication history, identifying and resolving medication discrepancies, creating a current medication list, and counselling on appropriate medication use. | Clinical pharmacists provided the intervention. They collaborated with the inpatient medical team to create the Best Possible Medication Discharge List. | The intervention was delivered individually and face‐to‐face. | |
| Pharmacist home visits | Older people often have trouble adhering to their medications. This study evaluated the effectiveness of a home‐based medication review on hospital admissions among elderly people. | A standardised visit form was used to record the home visit but no reference was provided. | Pharmacists arranged home visits with the participant during which they assessed the participant's ability to self‐medicate and drug adherence. They educated the participant, removed out‐of‐date drugs, reported drug reactions or interactions to the physician, and reported the need for a compliance aid. | Pharmacists conducted the home visits. Pharmacists held a postgraduate qualification in pharmacy practice or had recent continuing professional development in therapeutics. The pharmacists participated in a 2‐day training course, which included lectures on adverse drug reactions, prescribing in elderly people, improving concordance, and communication skills. | The intervention was delivered individually and face‐to‐face. | |
| Telephone consultation with home visits | Adherence to warfarin treatment and monitoring guidelines may be suboptimal among patients and staff. This study assessed the improvement in adherence to warfarin therapy with telephone and home visits. | A predesigned set of questions was used in the telephone consultation, but no reference or any additional details were provided. | The intervention group was counselled with once‐a‐week telephone consultations and 2 home visits per month by either a nurse or a pharmacist that dealt with warfarin use. | A pharmacist or a nurse provided the home visits. The telephone consultation was conducted by a pharmacist. | The intervention was delivered individually using a face‐to‐face format and telephone calls. | |
| Cardiovascular risk assessment and education sessions | Strategies for managing blood pressure are essential as high blood pressure is the leading risk factor for death. The study authors evaluated the effectiveness of a community‐based cardiovascular health promotion and disease prevention programme in reducing morbidity. | The Cardiovascular Health Awareness Program was a standardised intervention that consisted of 10 weeks of cardiovascular risk assessment, blood pressure measurements, and education sessions (CHAP 2017). | The intervention consisted of 10 weeks of cardiovascular risk factor assessment and educational sessions. Volunteers were recruited to help participants measure their blood pressure and supported self‐management by providing participants with their risk profile, risk‐specific educational materials and information about access to local services. At the end of the 10‐week programme and 6 months after the programme ended, the results were forwarded to family physicians who rank‐ordered their participants by their most recent systolic blood pressure reading. | Volunteers were recruited and trained to carry out the intervention. The volunteers were trained according to a standardised curriculum developed by a public health nurse and delivered by nurses working in the intervention community. | The intervention was conducted individually in a face‐to‐face manner. | |
| Educational intervention with pharmacist | Few studies have examined a pharmacist's contribution to improving diuretic compliance and reducing rehospitalisation and health care use. This study aimed to determine the impact of a pharmacist‐led intervention on patient‐guided diuretic dose adjustment. | The intervention group adjusted their diuretic dose using a flexible frusemide dose‐adjustment guide that was provided in the paper. | The intervention consisted of a 30‐min educational session and focused on improving participant self‐care, recognising symptoms of fluid retention, measuring weight daily, self‐adjusting the diuretic dose and improving knowledge of heart failure and heart failure medications. | A pharmacist provided the intervention. The frusemide dose‐adjustment guide was developed in collaboration with cardiologists. | Conducted individually in a face‐to‐face fashion. | |
| Pharmaceutical care plan | Regular medication reviews can reduce the risk of medication‐related problems. This study aimed to evaluate the effect of a pharmacist‐led medication review on pharmaceutical care issues and hospitalisations. | Clinically‐trained pharmacists completed a detailed profile for each participant using medical notes and computer records. All participants were interviewed in their home about their use of and responses to medication and their use of health and social services. No references provided | A pharmaceutical care plan was drawn up listing all pharmaceutical care issues together with all the actions planned to achieve the outcomes of any pharmaceutical care issue. Copies of the plan were given to the GP who was asked to agree, after which the pharmacist implemented the plan. | The pharmacist performed the medication review. The participants' GP indicated their level of agreement with each pharmaceutical care issue and with the actions taken. | The mode of delivery was individual and face‐to‐face. | |
| Use of health information technology to identify people at risk for delirium and falls, implement monitoring plans, and provide reports to pharmacists | Falls and delirium pose the greatest threats to resident safety in nursing homes and contributes to further functional decline. Medication use is associated with greater risk of delirium and falls. Therefore, this study used health information technology to identify residents at risk for delirium and falls due to adverse drug events. | A Geriatric Risk Assessment MedGuide was a database designed to identify medications that potentially contributed to delirium and fall risk (Tobias 1999). It also facilitated early recognition of signs and symptoms indicative of potential medication‐related problems. Training was provided to nursing staff and pharmacists in how to use the reports generated by the Geriatric Risk Assessment MedGuide. | Health information technology was used to identify residents at risk for delirium and falls, implement monitoring plans, and provide reports to pharmacists in conducting medication reviews. The consultant pharmacist shared the reports with the nurse contact at the facility and used the reports in their monthly drug review. | The intervention was an automated system that provided reports to pharmacists and nurses, who were trained to use these reports. The training for nurses provided information regarding medications that cause, aggravate, or contribute to the risk of falls and delirium. The course also reviewed symptoms and signs of adverse medication effects and reinforced the importance of the early observation of symptoms and signs of adverse medication effects. Pharmacists were trained to provide a targeted drug review for all participants who experienced delirium and falls. | The intervention was delivered individually and face‐to‐face. | |
| Home‐based medication review | Home‐based medication reviews are convenient for the patient and provide an opportunity to understand their medication‐taking in their home environment. Therefore, this study looked at whether home‐based medication reviews with elderly people could reduce hospital admissions. | The intervention comprised 2 home visits by a community pharmacist who educated the participant/carer about their medicines, noted any pharmaceutical care issues and assessed the need for an adherence aid. | At the home visit, the pharmacist educated the participant, removed out‐of‐date drugs, and assessed the need for an adherence aid. The pharmacist held regular meetings with the GP where changes to the participant's medications were discussed and amendments were implemented by the GP. | A pharmacist with a post‐graduate qualification in pharmacy practice conducted the home‐based medication review. They had regular meetings with the lead GP. Possible changes to the participant's medication were discussed and agreed amendments were implemented by the GP. | The intervention was delivered individually and face‐to‐face. | |
| Pharmacist medication review | Although angiotensin‐converting enzyme inhibitors and beta‐blockers reduce morbidity and mortality in people with heart failure, these treatments are underused. Pharmacists may improve treatment through medication review. This study investigated whether a pharmacist intervention would reduce hospital admission and death for people with heart problems. | Pharmacists received training covering the aetiology, symptoms, and evidence‐based management of heart failure. They also participated in monthly discussions of specific cases. The pharmacist used guidelines to optimise treatment for participants with left ventricular systolic dysfunction. All of these materials are available at onlinelibrary.wiley.com/journal | Participants were offered a 30‐min appointment with the pharmacist If there was agreement between the pharmacist and the participant, and subsequently with the doctor, medications were initiated, discontinued, or modified by the pharmacist during 3‐4 weekly or fortnightly consultations. | The pharmacists, who delivered the medication review, had between 3 and 16 years of post‐qualification experience, had experience delivering primary care‐based medication review clinics for people receiving multiple‐drug treatment and attended an in‐house training day covering the aetiology, symptoms, and evidence‐based management of heart failure. An additional session covered the methods of the trial. | The intervention was delivered individually and face‐to‐face. | |
| Pharmacist medication review | Aging and the use of polypharmacy are risk factors for drug‐related problems and medication‐related hospital admissions. Therefore, this study assessed the impact of a community pharmacist‐led medication review on hospital admissions in older people. | Pharmacists in the intervention group received a training course that covered the clinical management of older people and the medication review method. No reference was provided. | The medication review consisted of the pharmacist collecting information about the participant's health problems, medication use, lifestyle habits, and concerns about diseases and medications. The pharmacist then identified negative clinical outcomes related to medicines and drug‐related problems. Subsequently, an action plan was agreed upon which focused on participant outcomes and the medication use process. | Pharmacists provided the medication review. They received a 3‐day training course covering clinical management of elderly people, the medication review with follow‐up method, communication with participants and doctors, study protocol and documentation forms. | The intervention was delivered individually and face‐to‐face. | |
| Pharmacist visits | Pharmacists have adopted pharmaceutical care, which is the provision of drug therapy to improve a person's quality of life, to reduce morbidity and mortality. Unlike previous studies that did not focus on people who were most likely to benefit, this study examined veterans who were at high risk for a medication‐related problem. | Contacts between the pharmacist and participant were recorded on a data collection form, which contained the method of contact, time spent, medical problems addressed, drug‐related problems addressed, and drug‐related problems resolved. This form was not referenced. | The intervention participants received consultation and follow‐up care from a clinical pharmacist. | Pharmacists conducted the intervention. Most had a Doctor of Pharmacy degree and over 70% were either receiving or had completed postgraduate training. | The intervention was delivered individually and face‐to‐face. | |
| Home‐based nurse care | Home‐based nurse care can reduce adverse events in people with chronic heart failure. High levels of natriuretic peptides in people with heart failure are predictors of death and hospitalisations. The study authors looked at whether high levels of these peptides can predict whether people with heart failure benefit from a home‐based nurse intervention. | The nurse checked for and, in co‐ordination with the treating physician, implemented guideline‐based medication (Remme 1997; Remme 2001). | At home visits, the nurse checked and recorded weight, recorded symptoms and signs of heart failure as well as heart rate and blood pressure, and organised and reviewed blood analyses on demand. The nurse also gave the patient education and self‐management skills. | Nurses who specialised in caring for people with heart failure provided the intervention. | The intervention was delivered individually and face‐to‐face. | |
| Computerised care suggestions | Hypertension is associated with cardiovascular morbidity and mortality, but is difficult to control. Guidelines on hypertension are complicated and can become outdated quickly, so this study investigated the benefits of evidence‐based treatment for hypertension using a computerised system. | This study used the pharmacist intervention recording system, which was used to document all pharmaceutical care interventions (Overhage 1999). This system gave the pharmacist care suggestions, which they could pass on to the physician. The physician used an order writing workstation to write orders for drugs, tests, nursing activities, and consultations (McDonald 1999). The workstation gave the physician care suggestions for the treatment of hypertension. | The pharmacist intervention recording system was used by intervention pharmacists to receive care suggestions. The pharmacist could fill the prescription as written, discuss the suggestions with the participant and encourage discussions between the participant and physician, or contact the ordering physician. The physician intervention used an order‐writing workstation to write orders for drugs, tests, nursing activities and consultations and display care suggestions. All hypertension care suggestions were displayed as suggested orders along with possible actions and a brief explanation of the rationale for the suggestion. | Pharmacists and physicians provided the intervention. | The intervention was delivered individually and face‐to‐face. | |
| Home visits by physician assistants | Studies suggest that people who have undergone coronary artery bypass graft surgery benefit from a home intervention, but there are few studies of home visits by physicians or physician assistants. Therefore, this study examined the hospital readmissions of people who received home visits by physician assistants. | A physician assistant home care form/checklist was used to record all findings from the home visit. A copy of this form was provided in the paper. | Cardiothoracic physician assistants conducted home visits during which they performed a physical examination and reviewed the participant's medications. Adjustments were made to the participant's medications and new medications were prescribed as needed. The surgical wounds were examined and participant concerns were addressed. Prescriptions were written for antibiotics, blood work, or imaging studies. | Physician assistants provided the intervention. | The intervention was delivered individually and face‐to‐face. | |
| Pharmacist‐managed hypertension clinic | Hypertension can be controlled, but this study investigated whether it can be managed at a reasonable cost with minimal adverse effects by pharmacists. | Sitting blood pressure was measured with a Datascope Accutorr automated sphygmomanometer (Datascope Corporation Montvale, NJ, USA). 2 readings were taken for each participant and the average of the 2 readings was recorded (Datascope Patient Monitoring 1996). | Participants were counselled by a pharmacist who told them that efforts would be made to decrease the number of antihypertensive drugs or alter their therapy by giving more appropriate or less expensive drugs to achieve similar or improved blood pressure control. The pharmacist determined the most appropriate antihypertensive regimen for each participant, ordered laboratory tests as needed, and provided education on nonpharmacological ways to control blood pressure. | Clinical pharmacists provided the intervention. | The intervention was delivered individually and face‐to‐face. | |
| Pharmacist medication review | Pharmacists work with participants in designing, implementing and monitoring therapeutic plans, but elderly people may have problems with adhering to their medication. This study looked at treatment adherence, as well as hospitalisations and mortality, in elderly people who received a home visit by a pharmacist along with telephone follow‐up. | Pharmacists adhered to a manual to deliver the intervention (Medication Review‐Managing Medicine Manual, Danmarks Apotekerforening, Pharmakon. Medicingennemgang 2004). This manual helps pharmacists identify and resolve drug‐related problems (Danmarks 2004). | Participants were visited at home by a pharmacist who examined the medicines list with regard to side‐effects, interactions and administration. The pharmacist tried to make the regime less complex, informed participants, and motivated adherence. | Pharmacists who had some practical experience or courses in medication review provided the intervention. | The intervention was delivered individually. It was conducted by telephone and face‐to‐face. | |
| Pharmacist medication review | People with end‐stage renal disease take multiple drugs and experience multiple co morbidities, which places them at greater risk of drug‐related problems. This paper looked at the effects of a pharmacist‐led intervention on drug‐related problems and hospitalisations in ambulatory patients undergoing haemodialysis. | Drug‐related problems were recorded, evaluated and assigned to 10 possible categories (Hepler 1990). The drug‐related problems were also categorised into therapeutic drug classes and the outcome related to the drug‐related problem intervention was captured. | Participants assigned to pharmaceutical care had drug therapy reviews conducted by a nephrology‐trained pharmacist. The pharmacist conducted a participant interview, generated a drug therapy profile, identified and addressed drug‐related problems, and provided healthcare‐provider and participant education. The pharmacist also provided consultative services that focused on optimising drug therapy. | The clinical pharmacists who conducted the intervention were either nephrology‐trained or completing postdoctoral training in nephrology pharmacotherapy. | The intervention was delivered individually and face‐to‐face. | |
| Medication review, nurse education, and development of professional relationships | Pharmacist‐conducted medication reviews and nurse education about medication use may have an impact on drug use in nursing homes. This study looked at the effect of medication review and nurse education on mortality and hospitalisations in nursing homes. | Problem‐based educational sessions were provided to nurses and addressed basic geriatric pharmacology and some common problems in long‐term care. No referenced documentation is provided for these sessions. | The intervention introduced a new professional role to stakeholders with relationship building, nurse education, and a medication review by pharmacists. Professional contact between nursing home staff and pharmacists on issues such as drug policy and resident problems was conducted along with problem‐based educational sessions for nurses. These sessions addressed geriatric pharmacology and problems in long‐term care. The medication reviews highlighted adverse drug effects, ceasing or adding drugs, better use of specific drug therapy, non‐drug interventions, and adverse effect and drug response monitoring. | Clinical pharmacists delivered the intervention. | The intervention was delivered individually and in groups over the phone and face‐to‐face. | |
| Structured home visits by GP and nurse | Many hospital admissions are due to inappropriate medical treatment, and the discharge of fragile elderly patients is associated with a high risk of readmission. This study examined whether home visits by GPs and district nurses reduced the risk of readmission of discharged elderly patients. | The joint home visits were guided by an agenda. During the structured home visit the agenda included checking the discharge letter for recommended follow‐up, checking the need for adjustment of medication, checking if social and personal support was arranged, and checking the family’s medicine cabinet. This agenda was provided in the article. | There was a joint home visit by the GP and district nurse approximately one week after discharge from the hospital. 2 more contacts were conducted by the GP in the GP's clinic or as a home visit. These visits included checking the discharge letter, checking the need for adjustment of medication, checking if social and personal support was arranged, and checking the family's medicine cabinet. | GPs and district nurses provided the intervention. | The intervention was delivered individually and face‐to‐face. | |
| Pharmacist medication reviews | Adverse drug events are frequently caused by cardiovascular drugs. Pharmacists can identify and resolve drug‐related problems for people at home and reduce re‐hospitalisation rates. This study investigated whether a pharmacist‐led intervention could reduce re‐hospitalisations and death in people with heart failure. | Using a predefined checklist, the pharmacist tried to reduce the use of inappropriate mediations, encourage smoking cessation, suggest improvements in the participant’s diet, and promote medication adherence, self‐monitoring, and vaccination. The checklist is not provided in the paper. | The pharmacist in the intervention group conducted an in‐home medication assessment and 2 follow‐up visits. This involved assessing and reviewing physician notes and laboratory test values and interacting with prescribers on behalf of the participants. The pharmacist catalogued all medications and interviewed the participant regarding medication use. | A clinical pharmacist, who had over 20 years of combined experience as a hospital and community pharmacist and had received a doctor of pharmacy degree and completed a 1‐year clinical residency in home care, provided the intervention. | The intervention was delivered individually and face‐to‐face. | |
| Pharmacist medication review | Repeat prescribing is poorly managed in the UK, which puts people at risk. Pharmacists could review these prescriptions and reduce the pressure on GPs. This study tested whether pharmacists can review repeat prescriptions to reduce hospital admissions and deaths. | The process for reviewing repeat prescriptions involved discussing each condition with the participant and asking about symptoms (Lowe 2000). If clinical or pathological monitoring was due, the pharmacist directed the participant to the practice nurse or doctor. Participants with new clinical problems were referred to the doctor. | The pharmacists conducted a medication review during which they evaluated the therapeutic efficacy of each drug and the progress of the conditions being treated. Compliance, actual and potential adverse effects, interactions, and the participant’s understanding of the condition and its treatment were considered. The outcome of the review was a decision about the continuation of the treatment. | A pharmacist provided the medication review. | The intervention was delivered individually and face‐to‐face. | |
| Pharmacist medication review | Elderly people take multiple medicines, which increases the risk of adverse drug events. Pharmacists can improve medicine management for elderly people in the community. In this study, the authors looked at whether a pharmacist‐led review would reduce hospitalisations and deaths among elderly people in nursing homes. | The clinical medication review (Lowe 2000), which was conducted by the pharmacist, comprised a review of the GP clinical record, and a consultation with the participant and carer. The pharmacist made recommendations and passed them on a written proforma to the GP for acceptance and recommendation. | The pharmacist conducted a medication review in which the pharmacist identified the drugs that were taken, identified the original indication for each drug, assessed adherence to medication, and identified unaddressed medical problems. They also considered the continuing need for each drug, identified side effects, identified drug interactions or contraindications, and considered costs. Finally, the pharmacist implemented and documented any changes. | The study pharmacist provided the intervention. | The intervention was delivered individually and face‐to‐face. | |
| GP: general practitioner | ||||||
| Study | Location of intervention | When and how much of the intervention was delivered | Tailoring | Modifications | Adherence planning | Adherence assessment |
|---|---|---|---|---|---|---|
| 83 community pharmacies in the provinces of Asturias, Barcelona, Madrid and Biscay | The intervention was delivered once. | There was no tailoring made to the intervention. | Two additional seminars were given to the intervention group on real cases in order to approve the intervention. | Not undertaken | Not undertaken | |
| Community pharmacies in 7 European countries; Denmark, Germany, The Netherlands, Northern Ireland (co‐ordinating centre), Portugal, Republic of Ireland and Sweden. A minimum of 12 sites per country were chosen according to specific criteria set within each participating country relating to the population of elderly people who visited the pharmacy, staffing levels within the pharmacy and working relationships with local GPs. | The intervention was delivered at least once according to the study manual. However, Each site was free to provide as much information as possible to the intervention group as per the study manual. | A study manual describing the intervention was developed for all the participating countries. Each country translated the manual into their own language. | Each country adapted the manual, translating and modifying sections where appropriate, according to differing national practices. | Not undertaken | Not undertaken | |
| 7 Primary Health care clinics in Mataró and Argentona | The intervention included 3 phases and the participants were followed up for 12 months. It is not clear if the intervention was repeated more than once. | There was no tailoring made to the intervention. | There were no modifications made to the intervention during the study. | Not undertaken | Not undertaken | |
| 9 primary care physician practices in Washington State. Clinics were allowed to select their target condition of focus: frail older adults or people with diabetes. The physicians were board certified in Family Practice and did not have formal training or certification in geriatric medicine. | The intervention was undertaken once. However there was variability in the frequency of one of its components. | There was no tailoring made to the intervention. | There were no modifications made to the intervention during the study. | A priori process of care measures for each of the geriatric syndromes were developed with decision rules for acceptable documentation by the study reviewers for the interventions. | The chart abstraction of assessing the documentation for the interventions was performed by one member of the study team along with an additional reviewer blinded to knowledge of the study group and study hypothesis. The overall level of agreement between the 2 reviewers was acceptable based on published ranges (kappas for geriatric syndrome process measures 0.75 to 0.85) | |
| Chronic care geriatric facilities in Central Israel | The intervention was done once at 6 months and 12 months later. | There was no tailoring made to the intervention. | There were no modifications made to the intervention during the study. | Not undertaken | Not undertaken | |
| A private organisation of 37 nursing homes in Spain | It is unclear how many times the intervention was given as the educator offered further on‐demand advice on prescription for the next 6 months. | There was no tailoring made to the intervention. | There was no modifications made to the intervention during the study. | Not undertaken | Not undertaken | |
| Home health patients within a medicare insured home health population in Canada | The intervention was undertaken at least once however, some participants received more than one phone call as additional telephone follow‐up was provided as needed per the pharmacists' discretion during the first 30 days of the 60‐day home healthcare episode. | Some participants received additional follow‐up depending on their conditions. | There were no modifications made to the intervention during the study. | Not undertaken | Not undertaken | |
| Large multispecialty group practice employing 265 physicians, including 66 primary care providers caring for adults in the outpatient setting | Daily records generated by the computer system were examined. | There was no tailoring made to the intervention. | There were no modifications made to the intervention during the study. | Not undertaken | Not undertaken | |
| 804‐bed academic medical centre in North Carolina, USA | The intervention took place once. | There was no tailoring made to the intervention. | Only hospitalisations and ED visits at the study institution were included for those participants who were not able to be contacted after 3 phone call attempts. | Not undertaken | Not undertaken | |
| Home‐based medication review after discharge from acute or community hospitals in Norfolk and Suffolk, UK. | The intervention was performed once. | It is possible that a small number of participants in both groups may have had their medication reviewed during the follow‐up period by their GP or community pharmacist. | There were no modifications made to the intervention during the study. | Not undertaken. No data on adherence were collected. | Not undertaken | |
| Telephone consultation with home visits | The intervention was performed once. | Any additional contact as requested by the participant in the intervention group was undertaken. | There were no modifications made to the intervention during the study. | Not undertaken | Not undertaken | |
| Community‐based pharmacies in Canada | The intervention was performed once as planned. | The local lead organisations used several strategies to recruit volunteer peer health educators. These strategies included using the local lead organisation's existing volunteer base, advertising in the local media, and giving presentations at local seniors’ clubs. | When required, Cardiovascular Health Awareness Program support staff produced and mailed invitation letters on behalf of participating physicians (CHAP 2017). | Feedback of results was given to primary healthcare providers. | Evaluation data collected for the purpose of ongoing evaluation and quality improvement: 1. Success of different advertising/invitation strategies 2. Attendance, consent, completed assessments 3. Nurse assessments, pharmacist consults, fax/call to family physician the same day. Feedback to family physicians, pharmacists, and participants | |
| Outpatients clinic in Melbourne, Australia | The intervention was performed once as planned. | There was no tailoring made to the intervention. | There were no modifications made to the intervention during the study. | There were written instructions on how to adjust the dose of frusemide per weight increase. | Data on dosage adjustment of frusemide were collected and compared against the initial criteria. | |
| General medical practices in the Grampian region of Scotland | The intervention was performed once as planned. | In the control group, when pharmacists considered a review to be serious and beneficial to the participants, an independent medical assessor decided on the need to withdraw the participants on clinical grounds. | There were no modifications made to the intervention during the study. | Any outstanding care issues in both groups were communicated to the participant's GP. | Not undertaken | |
| 25 nursing homes serviced by 2 long‐term care pharmacies in Northern Ireland | It is unclear the number of times the reports were generated and used by the pharmacists for every resident. | The Geriatric Risk Assessment MedGuide database software for falls and delirium was integrated into the pharmacies’ commercial pharmacy software system (Rescot LTCP System) for the intervention homes (Tobias 1999). | It is unclear if there were any modifications to the interventions. | The computer system did not capture if the recommendations done by the pharmacist were accepted. | Not undertaken | |
| A GP setting in Norfolk, UK | It is unclear how many times the pharmacist and the GP met to discuss participant's care plan. | A follow‐up visit with the participant occurred 6‐8 weeks later to reinforce the original advice, and assess whether there were any further pharmaceutical care issues to address with the GP. | It is unclear if there were any modifications to the interventions. | Not undertaken | Not undertaken | |
| The study was conducted within the NHS which provides free health care to the population of the UK. 27 primary care‐based pharmacists employed by the NHS to work with family doctors | It is unclear how many times the pharmacist met the participant and the GP. | If there was agreement between the pharmacist and the participant during the consultation and subsequently with the family doctor, medications were initiated, discontinued, or modified by the pharmacist during 3‐4 subsequent weekly or fortnightly consultations. | It is unclear if there were any modifications to the interventions. | Not undertaken | Not undertaken | |
| The study was conducted in 178 community pharmacies in Spain | It is unclear how many times the intervention was undertaken. | A specifically trained pharmacist called a practice change facilitator helped pharmacists of the intervention group in the provision of the medication review with follow‐up service, identifying barriers specific to each pharmacy and providing solutions. | It is unclear if there were any modifications to the interventions. | The practice change facilitator ensured fidelity to the intervention and supported pharmacists of both study groups on queries about documentation forms. | The experts were requested to answer individually for each case and the degree of agreement between them was later established. Inter‐rater reliability was measured using Fleiss's kappa. | |
| 9 Veterans Affairs medical centres in the USA | It is unclear how many times participants were seen by the pharmacist in the intervention group as the protocol indicated that each participant should have at least 3 visits with the clinical pharmacist during the study, but participants could be seen as frequently as deemed necessary to ensure appropriate care. | To prevent contamination, some sites marked medical records of intervention and control participants to alert clinical pharmacists that participants were in the study. Other sites noted this distinction in electronic medical records. One site distributed a list of participants enrolled in the study to all pharmacists providing primary care. | Clinical pharmacist intervention, however, occurred in one control participant; this participant was withdrawn from the study and his data were not included in the results. | Each contact with the participant was recorded on a standard data collection form that contained information about the method of contact, estimated time spent with the participant, medical problems addressed, drug‐related problems addressed, and drug‐related problems resolved. | Each month after enrolment the co‐ordinating centre received electronic data on each participant's prescription drugs dispensed in the preceding month. When participants either completed the study or died, data on resource use from enrolment to termination were retrieved. | |
| Ambulatory patients participating in the EuroHeart Failure Survey programme in Vienna | It is unclear how many times the nurse visited the intervention participants as more visits were made optional for participants. | More frequent contacts such as visits or telephone calls between the nurse and the participants were optional in case the participant or the nurse considered them necessary. | The nurse was in charge of individualised participant and caregiver education and enhancement of self‐management. If the nurse noted any deterioration in the participant's status, she reported to the treating physician or advised the participant to visit the treating physician. | Not undertaken | Not undertaken | |
| Academic primary care internal medicine practice in the USA | It is unclear how many times the intervention was undertaken. | There was no tailoring made to the intervention. | There were no modifications made to the interventions. | Data necessary to generate care suggestions were derived from the computer programme. Treatment suggestions fell into 5 major categories. | Not undertaken | |
| Ambulatory patients discharged from a large 702‐bed hospital in Staten University Hospital, USA | It is unclear how many times the physician visited each participant in the home after their discharge. | There was no tailoring made to the intervention. | There were some modifications done to the intervention due to the participants not being available at the weekend. Participants were not seen directly after discharge as per the study protocol. | All findings were documented on the intervention visit form. | It is unclear if this was undertaken. | |
| Managed care organisation in California, USA | It is unclear how many times participants were seen by the pharmacist in the intervention group as additional follow‐up was organised by the pharmacists for some participants. | Additional follow‐up was organised by the pharmacists for some participants. | The intervention was not modified. | Not undertaken | Not undertaken | |
| Patients living at home in the municipality of Aarhus, Denmark | The intervention was performed at the intended follow‐up. | Pharmacists could consult the project physician if they considered a participant's medication problems to be life‐threatening. | The intervention was not modified. | Adherence to the medications were assessed by a pill‐count in all participants during 1 year. | Pill count was undertaken | |
| The study took place in a non‐profit university‐affiliated dialysis clinic in Albany, USA. | It is unclear if all participants received the same number of follow‐up visits by the pharmacist or the physician in the intervention group. | It is unclear if there was any tailoring made to the intervention. | The intervention was not modified. | Not undertaken | Not undertaken | |
| 52 nursing homes located in south‐east Queensland and north‐east New South Wales, Australia | There was variability in the number of educational sessions provided to staff in each nursing home as well as the number of visits by the intervention pharmacists. | It is unclear if there was any tailoring made to the intervention. | It is unclear if the intervention was modified. | Validation of prescription claim data with participants' medications profiles. | To validate prescription claims data, a sample of 1328 cross‐sectional medication profiles were collected for 8 nursing home clusters for control and intervention homes at post‐intervention. An audit, comparing original post‐intervention medication data with the same data recollected up to 6 weeks later for a 6%, random sample, showed an overall reproducibility of 97% (range 92% to 100%) | |
| Patients discharged from Glostrup Hospital, Denmark. | The intervention was performed as prescribed. | There was no tailoring made to the intervention. | There was no modification made to the intervention. | Not undertaken | Not undertaken | |
| Heart failure patients discharged from hospital in Albany, Scotland | The intervention was performed as prescribed. | The clinical pharmacist accessed and reviewed all pertinent physician notes and laboratory test values via the National Endowment for the Humanities data system and interacted with prescribers on behalf of the participants, as necessary. | There was no modification made to the intervention. | Not undertaken | Not undertaken | |
| 4 GPs in Leeds, UK | It is unclear how many times the pharmacist visited the participant. | Immobile participants were visited at home. Non‐attenders were invited once more by telephone. | The study authors agreed with each practice the level of intervention that the pharmacist could make without seeking prior approval | It is unclear if this was undertaken. | It is unclear if this was undertaken. | |
| 65 care homes for the elderly in Leeds, UK | It is unclear how many times the pharmacist reviewed each participant. | There was no tailoring made to the intervention. | There was no modification made to the intervention. | Pharmacists filled in a proforma sheet including their recommendations. | GP acceptance of the recommendations was signified by ticking a box on the proforma. | |
| ED: emergency department; GP: general practitioner; NHS: National Health Service | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1.1 Number of hospital admissions Show forest plot | 2 | 3889 | Risk Ratio (M‐H, Random, 95% CI) | 1.24 [0.79, 1.96] |
| 1.2 Number of people admitted to hospital Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 1.3 Number of emergency department visits Show forest plot | 2 | 1067 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.71 [0.50, 1.02] |
| 1.4 Mortality Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 2.1 Number of hospital admissions Show forest plot | 11 | 6203 | Risk Ratio (M‐H, Random, 95% CI) | 0.85 [0.71, 1.03] |
| 2.2 Number of people admitted to hospital Show forest plot | 13 | 152237 | Risk Ratio (M‐H, Random, 95% CI) | 0.92 [0.86, 0.99] |
| 2.3 Number of emergency department visits Show forest plot | 5 | 1819 | Risk Ratio (M‐H, Random, 95% CI) | 0.75 [0.49, 1.15] |
| 2.4 Mortality Show forest plot | 12 | 154962 | Risk Ratio (M‐H, Random, 95% CI) | 0.94 [0.85, 1.03] |
